vEPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/540/R-92/001
January 1992
Technology Evaluation
Report
Ogden Circulating Bed
Combustor at the
McColl Superfund Site
-------
-------
EPA/540/R-92/001
January 1992
TECHNOLOGY EVALUATION REPORT
OGDEN CIRCULATING BED COMBUSTOR
at the
McCoIl Superfund Site
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Printed on Recycled Paper
-------
NOTICE
The information in this document has been funded, wholly or in part, by the U.S. Environmental
Protection Agency under Contract No. 68-03-3255 to Foster Wheeler Enviresponse, Incorporated and the
Superfund Innovative Technology Evaluation (SITE) Program. It has been subjected to the Agency's peer and
administrative review, and has been authorized for publication as an EPA document. Mention of trade names
or commercial products does not constitute an endorsement or recommendation for use.
n
-------
FOREWORD
The Superfund Innovative Technology Evaluation (SITE) Program was authorized! in the 1986 Superfund
amendments. The Program is a joint effort between USEPA's Office of Research and Development and the
Office of Solid Waste and Emergency Response. The purpose of the Program is to assist the research,
development, and demonstration of hazardous waste treatment technologies necessary to implement new cleanup
standards which require greater reliance on permanent remedies. This is accomplished through technology
demonstrations which are designed to provide engineering and cost data on selected technologies. This treatability
study will determine if a SITE Demonstration will occur.
This report provides documentation of a pilot-scale treatability study and demonstration of an innovative
technology that took place in San Diego, California at the Ogden Environmental Services (OES) Research facility.
Observation and sampling of a research-scale circulating bed combustor (CBC) took place during a preliminary
test sponsored under an agreement between the SITE Program and OES. The treatability study was conducted
on a minimal amount of representative McColl Site waste to determine if a full SITE Demonstration is feasible.
E. Timothy Oppelt
Director
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
in
-------
ABSTRACT
An evaluation of the Ogden Environmental Services (OES) circulating bed combustor (CBC) technology
was carried out under the Superfund Innovative Technology Evaluation (SITE) Program to determine its
applicability as an on-site treatment method for waste site cleanups, and more specifically for use at the McColl
Superfund Site in Fullerton, California, as a candidate technology to be tested in the SITE Program.
A treatability study and demonstration of the technology was performed at the OES research facility in
San Diego, California in March 1989. Operational monitoring, process stream, and flue gas sampling and
analysis data were collected and carefully evaluated to determine the technology's capability to destroy McColl
Site contaminants and permit operation of a transportable, field-erectable unit at the McColl Site.
The report includes an Executive Summary presenting the program evaluation criteria, descriptions of
the McColl Site and the CBC process, program results, and conclusions; a description of the SITE Program and
the program performance criteria; the vendor's description of the CBC process; the test conditions from waste
feed preparation through the sampling and analysis program; an evaluation of the treatability study results; a
listing of the cost components of the study; and an encapsulation of the test results and conclusions.
The test results indicate that continued deliberations among the federal, state, and local agencies, and
discussions with the technology developer are necessary to establish the conditions to conduct an on-site
Demonstration test.
IV
-------
CONTENTS
Page
Notice u
Foreword "i
Abstract iv
Contents v
Abbreviations and Symbols vii
Acknowledgment x
Executive Summary 1
Introduction : 1
Site Description 2
Vendor's System Description 2
Sampling and Analysis Program 4
Treatability Study Results 4
Conclusions 5
Introduction 7
The SITE Program 7
McColl Site Description 7
Program Objective 9
Performance Criteria 11
Key Contacts 12
Vendor's System Description 13
Introduction 13
System Description 13
Process Control 15
Process Monitoring and Data Acquisition 17
Test Program Procedures 19
Waste Feed Preparation 19
Operating Conditions and Operating Range 19
Sampling and Analysis Program 20
QA/QC Plan and Audits 27
Performance Data Evaluation 33
Introduction 33
Performance Criteria Results 34
Quality Assurance/Quality Control 45
Treatability Study Costs 51
Conclusions 53
Test Results 53
Recommendations 57
References
59
-------
Number
Figure 1.
Figure 2.
Figure 3.
Figure 4.
FIGURES
Schematic flow diagram of CBC for waste treatment 3
McColl Site layout 8
Schematic of the OES CBC system 14
Sampling and monitoring points of the CBC system 24
TABLES
Number Page
Table 1. Chemical Characterization of McColl Waste 10
Table 2. Baghouse Filter System 16
Table 3. Interlock Settings 17
Table 4. Test Durations 20
Table 5. Treatability Study Process/Permit Conditions 21
Table 6. Description of CBC Monitoring Parameters and Methods 22
Table 7. Process Stream Sampling and Analysis 23
Table 8. Flue Gas Sampling and Analysis 25
Table 9. Flue Gas Analyzers 27
Table 10. List of Volatile Organics 28
Table 11. HSL Semivolatile Organics 29
Table 12. TCLP List of Volarile Organics 30
Table 13. Number and Types of QC Samples 31
Table 14. Analytical QC Measures 32
Table 15. DREs for CC14 34
Table 16. Combustion Emission Concentrations for Selected Compounds 35
Table 17. Flue Gas Results 36
Table 18. Chloride in the Flue Gas 37
Table 19. Continuous Emissions Monitoring Results 39
Table 20. Organics and Halide Results 40
Table 21. Total Metals Results , 42
Table 22. TCLP Results 43
Table 23. Summary of Physical Parameters 44
Table 24. Percent Recoveries Results 46
Table 25. Treatability Study Costs : 52
Table 26. Comparison of CBC Parameters to Other Tests 54
VI
-------
ABBREVIATIONS AND SYMBOLS
acftn actual cubic feet per minute
AEERL Air and Energy Engineering Research Laboratory
APCD Air Pollution Control District
ATC Alliance Technologies Corporation
Btu British thermal unit
BTEX benzene, toluene, ethylbenzene, xylene
Ca calcium
CBC circulating bed combustor
CC14 carbon tetrachloride
CEM continuous emissions monitor
CERCLA Comprehensive Environmental Response, Compensation, and Liability Act
cftn cubic feet per minute
CO carbon monoxide
CO2 carbon dioxide
CRT cathode ray tube
cu cubic
DAS data acquisition system
DHS Department of Health Services
DOT Department of Transportation
DP differential pressure
DRE destruction and removal efficiency
dscfrn dry standard cubic ft/min
dsl dry standard liter
EIR environmental impact report
EPA Environmental Protection Agency
FGC flue gas cooler
ft feet
FWEI Foster Wheeler Enviresponse, Inc.
gal gallon
GC/MS gas chromatography/mass spectrometry
gpm gallons per minute
gr grain
HC1 hydrochloric acid
HEPA high efficiency particulate air (filter)
hr hour
HSL hazardous substances list
1C ion chromatography
ICAP inductively coupled argon plasma
ID induced draft
i.d. inside diameter
in. inch
ITL Industrial Testing Laboratories
kg kilogram
L liter
Ib pound
LCS laboratory control samples
mg milligram
min minute
vn
-------
ABBREVIATIONS AND SYMBOLS (continued)
ml
MM
MM5
NBS
ng
NIH
NOX
NPL
02
OES
ORD
OSWER
PCBs
PCDD
PCDF
PIC
PID
PLC
POHC
ppm
psig
QA/QC
QAPP
RCRA
RD&D
RI/FS
RPD
RREL
RTI
sec
S
S&A
SARA
SCAQMD
scfm
SITE
SO2
SOX
sq
SROA
SV
TCL?
THC
TIC
TLI
TOC
ug
USEPA
milliliter
million
modified method 5
National Bureau of Standards
nanogram
National Institutes of Health
nitrogen oxides
National Priority List
oxygen
Ogden Environmental Services, Inc.
Office of Research and Development
Office of Solid Waste and Emergency Response
polychlorinated biphenyls
polychlorinated dibenzo-p-dioxin
polychlorinated dibenzo furan
product of incomplete combustion
proportional integral derivative
programmable logic controller
principal organic hazardous constituents
parts per million
pounds per square inch (gauge)
quality assurance/quality control
quality assurance project plan
Resource Conservation and Recovery Act
research, development, and demonstration
remedial investigation/feasibility study
relative percent difference
Risk Reduction Engineering Laboratory
Research Triangle Institute
second
sulfur
sampling and analytical
Superfund Amendments and Reauthorization Act of 1986
South Coast Air Quality Management District
standard cubic feet per minute
Superfund Innovative Technology Evaluation
sulfur dioxide
sulfur oxides
square
supplemental revaluation of alternatives
semivolatile
toxicity characteristic leaching procedure
total hydrocarbon
tentatively identified compound
Triangle Laboratories, Inc.
total organic carbon
microgram
United States Environmental Protection Agency
vin
-------
V
VOHC
VOST
wg
ABBREVIATIONS AND SYMBOLS (continued)
volume
volatile organic hydrocarbon
volatile organics sampling train
water gauge
IX
-------
ACKNOWLEDGMENT
This report was prepared under the direction and coordination of Douglas W. Grosse, EPA SITE Manager
in the Risk Reduction Engineering Laboratory, Cincinnati, Ohio. An extensive contribution in both project
coordination and documentation review was provided by John Blevins of EPA Region IX. Valuable assistance
was also given by EPA's Office of Research and Development; the California Department of Health Services
(DHS); Brian Baxter, Robert Wilbourn, and the operating staff of Ogden Environmental Services, Inc.; S-CUBED
(audit services); and Alliance Technologies Corporation (sampling and analytical services).
This document was prepared for EPA's Superfund Innovative Technology Evaluation Program by G.W.
Sudell of Foster Wheeler Enviresponse, Inc. under Contract No. 68-03-3255.
-------
SECTION 1
EXECUTIVE SUMMARY
INTRODUCTION
The SITE Program treatability study and demonstration of the Ogden Environmental Services (OES)
circulating bed combustor (CBC) was conducted during March 1989 at OES's research facility in San Diego,
California. The main objective of the study was to evaluate the effectiveness of this technology in thermal
destruction of hazardous wastes in feedstock from the National Priority List (NPL) McColl Superfund Site in
Fullerton, California and, thus, to determine if the technology is a viable candidate for further on-site testing
under the SITE Program. OES currently operates two transportable, field-erectable, 100-tori/day (10 MM Btu/hr)
CBC units and has two additional units under fabrication.
EPA conducted the demonstration from March 27 to March 31,1989, and processed approximately 8,600
Ib of material, of which 5,500 Ib were McColl waste and contaminated soil. The nominal 2 million Btu/hr
pilot-scale CBC can process approximately 1,000 Ib/hr of waste feed. It was limited during the treatability study,
by permit, to less than 500 Ib/hr of waste feed. During this period the EPA SITE staff, EPA Region IX, DHS,
San Diego Air Pollution Control District (APCD), and EPA contractors observed the CBC operation, collected
data, and documented the test proceedings.
To evaluate the CBC these principal performance criteria were applied:
DREs - the CBC's calculated destruction and removal efficiency for organic compounds.
o
o
Flue gas emissions — the ability of the unit and its associated air pollution control system to control acid
gas emissions, particulate matter, and criteria air pollutants.
o Organics destruction — the ability of the system to destroy hazardous organic constituents present in the
feed, and the effectiveness of the CBC system in minimizing undesired combustion by-products and
products of incomplete combustion (PICs).
o Toxic metals distribution — determining the partitioning of metals into the fly ash, bed ash, and flue gas
streams.
o TCLP results — evaluating whether or not toxic metal and organic contaminants in the waste feed could
be rendered non-leachable in the ash residue thereby making the ash non-hazardous.
o Physical parameters — evaluating the physical/chemical changes of the waste feed after processing.
o System operating conditions — effectiveness of current instrumentation and controls, including the ability
to control critical process parameters and to respond to transient and upset conditions; and the
effectiveness of sulfur dioxide (SO2) capture and the calcium-to-sulfur (Ca/S) ratios needed to obtain this
effectiveness.
o QA/QC — determining the quality of the collected data.
The demonstration was designed to collect as much data as possible to define the system operation. In
addition to the information areas above, the following were to be collected: material balances based on the unit
operations and material flows; mechanical operations history of the unit; operational requirements and system
reliability; operating and maintenance data including labor, energy, and supply requirements, scheduled
1
-------
maintenance requirements, equipment durability and reliability, and operational safety; pretreatment and
posttreatment requirements to enable on-site operation within existing regulatory limits; unit costs to effectively
develop a cost/economic analysis for the unit; and energy effectiveness.
SITE DESCRIPTION
The McColl Superfund Site was used from 1942-1946 for the disposal of acidic sludges resulting from
the alkylation and product-treating processes used in the refining of aviation gasoline. The waste is characterized
as having a low pH (approximately 2); a high sulfur content — making it quite malodorous; elevated
concentrations of organic sulfur, aromatics, and aliphatic hydrocarbons; and low levels of some toxic metals such
as chromium, arsenic, and cadmium.
In 1983 the California Department of Health Services (DHS) completed a remedial investigation/feasibility
study (RI/FS) to determine the best course of action to be taken to remediate the site. In 1984, the EPA and DHS
decided on an optimum course: excavation, removal, and transportation of the waste to an off-site facility for
final disposal. Execution of the remedy was halted in 1985 by a state court injunction requiring the preparation
of an environmental impact report (EIR) prior to removal of material from the site. In 1989 EPA completed a
supplemental reevaluation of alternatives (SROA), which concluded that thermal destruction was a viable method
for use in the site remediation. To take advantage of a full-scale on-site field demonstration, EPA began
negotiations with the McColl residents and local authorities to gain their support for an EPA-sponsored on-site
Demonstration of the transportable OES CBC system. The McColl residents and the McColl Interagency
Committee asked EPA to conduct a treatability study and demonstration at the OES research facility to provide
data showing the effectiveness of treating the McColl waste before bringing a full-scale unit onto the McColl Site.
VENDOR'S SYSTEM DESCRIPTION
The information in this section on the CBC system has been provided, reviewed, and approved by Ogden
Environmental Services, Inc. The EPA makes no claims as to the accuracy of the statements.
The CBC is a thermal destruction system that uses high-velocity air to entrain circulating solids in a
highly turbulent combustion zone (Figure 1). This design allows combustion along the entire length of the
reaction zone. Solids, slurries, or liquids can be introduced into the combustor loop where they contact hot bed
material recirculating through the cyclone. When introduced into the primary combustion zone, the waste heats
rapidly and continues to be exposed to high temperatures (up to 1800°F) throughout its residence time
(approximately 1.5 sec for gases) in the combustion loop. High velocity air entrains the circulating soil, which
travels upward through the combustor and into the cyclone. The cyclone separates the combustion gases from
the hot solids. The solids then are returned to the combustion chamber via a proprietary non-mechanical seal.
As a consequence of the highly turbulent combustion zone, temperatures around the entire combustion loop are
uniform to within ±50°F. The hot flue gases and fly ash pass through a convective flue gas cooler into a
baghouse filter which traps the ash. Filtered flue gas then exits to the atmosphere. Heavier particles of purified
soil remaining in the combustor lower bed are removed slowly by a water-cooled bed ash conveyor system.
Acid gases and sulfur oxides formed during combustion are captured by limestone added directly into the
combustor. The reaction of limestone and hydrochloric acid (HC1) forms calcium chloride; the reaction of
limestone and SOX forms calcium sulfates. The combustion and neutralization of the acid gases within the
combustion chamber eliminate the need for afterburners and add-on scrubbers. Emissions of CO and NOX are
controlled to low levels by the turbulent mixing, low temperatures (1,425°F to 1,800°F), and staged combustion
achieved by injecting secondary air at sequenced locations in the combustor. OES has found that because of the
high heat of combustion and turbulent mixing, the CBC is capable of attaining required DREs for both hazardous
wastes (99.99%) and toxic wastes (99.9999%) at temperatures below those used in other types of incinerators
(typically >2,000°F) [2,3].
-------
SOLID
WASTE
COMBUSTION AIR
Figure 1. Schematic flow diagram of CBC for waste treatment.
3
-------
The CBC technology accomplishes pollution control through a high efficiency fabric baghouse filter to
control particulate releases; low operating temperatures, minimum excess air, and staged combustion to control
NOX emissions; gas/solids-mixing to control CO and volatile organic hydrocarbon (VOHC) emissions; and the
addition of limestone to the combustor for in situ scrubbing of acid gases, with no liquid waste streams.
SAMPLING AND ANALYSIS PROGRAM
EPA contracted the services of the Alliance Technologies Corporation laboratories to sample and evaluate
the physical and chemical characteristics of the waste feed, bed ash, and fly ash for volatile and semivolatile
organics, metals, dioxins and fiirans, polychlorinated biphenyls (PCBs), and organic and inorganic halogens; the
limestone feed for metals and calcium; and the stack gas stream for volatile and semivolatile organics, particulates
and chlorides. OES collected operational data and conducted continuous emissions monitoring of the stack gas
for SO2, CO, NO, oxygen, CO2, and THC.
TREATABILITY STUDY RESULTS
Destruction and Removal Efficiency
DREs for the CC14 performance indicator met the 99.99% limit established by RCRA for trial burns.
Flue Gas Emissions
Acid Gas Removal ~
No (total) chloride was detected in the flue gas. The chloride emission rate (based on minimum detectable
limits) was calculated to be less than 0.0083 Ib/hr, well below the RCRA performance standard of 4 Ib/hr.
Flue Gas Particulates —
Particulate concentrations in the flue gas averaged 0.0029 gr/dscf at 7% O2, as compared to the RCRA
limitation of 0.08 gr/dscf at 7% O2. Although this result is lower than normal permitted emission levels, the
California South Coast Air Quality Management District (SCAQMD) requires that particulate emissions not
exceed 0.002 gr/dscf corrected to 12% CO2 (equates to approximately 0.0022 gr/dscf at 7% O2). Further testing
of the CBC would be necessary to ensure compliance to SCAQMD emission limits.
Flue Gas Organics and Metals —
Only low concentrations of hazardous organics and metals were detected in the flue gas.
CEM Results -
CO, NOX, and THC emissions were controlled below the OES facility permit limits of 250 ppm for CO
and 100 ppm for THC. (NOX emissions were not permit-specified.) SO2 results were not quantifiable because,
during the study, the SO2 analyzer was not operating within performance specifications.
Organics Destruction
Organics originally present in the waste feed were not detected in the fly ash and bed ash streams, and
were low in the flue gas. PCBs, PCDDs, and PCDFs were not detected in the waste feed, residual ash, or flue
gas streams. PIC concentrations and concentrations of other undesired combustion by-products appear to have
been minimized in this study, and will be further evaluated in the SITE test. Organic chloride detected in the
Run 3 sample was attributed to the CC14 used to spike the sample.
-------
Toxic Metals Distribution
Metals in the waste feed were partitioned into the ash and flue gas streams. Ash leachate tests defined
the ash residue as non-hazardous. The effect of the metals concentrations in the flue gas will be evaluated in
health risk assessment for the on-site Demonstration Test.
TCLP Results
Toxicity characteristic leaching procedure (TCLP) test results on the ash stream were below the
established TCLP regulatory limits. However, the high pH of the ash matrix may have affected the results. No
significant organic concentrations were found in the leachate.
Physical Characteristics
No anomalies were found in the physical/chemical results.
System Operating Conditions
Except for initial feed interruptions caused by the SO2 and THC interlocks operating, the system
performed reliably. However, the relatively high heating value and variable SO2 content of the McColl feedstock
will result in transient surges that will affect the operation of the CBC. The effectiveness of the CBC control
system in reacting to intentionally-induced abnormal operating conditions — intended to test its reaction to feed
variations — was not determined because of the facility permit conditions. The optimum Ca/S ratio required to
limit SO2 emissions also could not be evaluated because of the permit limitations.
Quality Assurance/Quality Control
The QA/QC analytical results indicated that reliable data was obtained from the study (except for the SO2
analyzer data, as reported in the audit results).
Costs
OES's total costs for the demonstration were calculated to be $193,505 based on permitting, labor,
supplies, facility modifications, and decontamination. EPA costs were $620,000 for waste preparation,
transportation, sampling and analysis, waste disposal, and report preparation. More detailed costs will be
developed during the SITE Demonstration Test at the McColl Site.
CONCLUSIONS
It is concluded that the CBC technology can be evaluated further under the SITE Program. The SITE
Demonstration Test should be designed to test the CBC at its full operating capacity. The viability of this option
is dependent upon further deliberations among the regulatory agencies, discussions with the technology developer,
and continued evaluation of the treatability study results and other available test data to verify that the CBC can
meet the regulatory standards.
-------
-------
SECTION 2
INTRODUCTION
THE SITE PROGRAM
In 1986, the EPA's Office of Solid Waste and Emergency Response (OSWER) Jind Office of Research
and Development (ORD) established the Superfund Innovative Technology Evaluation (SITE) Program to promote
the development and use of innovative technologies to clean up Superfund sites across the country. Now in its
fifth year, the SITE Program is helping to provide treatment technologies necessary to implement new federal
and state cleanup standards aimed at permanent remedies, rather than quick fixes. The SITE Program is
composed of three major elements: the Demonstration Program, the Emerging Technology Program, and the
Monitoring and Measurement Technologies Program.
The major focus has been on the Demonstration Program, which is designed to provide engineering and
cost data on selected technologies. To date, the demonstration projects have not involved funding for technology
developers. EPA and developers participating in the program share the cost of the demonstration. The
developers are responsible for demonstrating their innovative systems at chosen sites, usually Superfund sites,
while EPA is responsible for sampling, analyzing, and evaluating all test results. The result is an assessment of
the technology's performance, reliability, and cost. This information is used in conjunction with other data to
select the most appropriate technologies for the cleanup of Superfund sites.
Developers of innovative technologies apply to participate in the Demonstration Program by responding
to EPA's annual solicitations. EPA also will accept proposals at any time when a developer schedules a project
to treat Superfund waste. To qualify for the program, a new technology must be at the pilot- or full-scale
development stage and offer some advantage over existing technologies. Mobile technologies are of particular
interest to EPA.
Once EPA has accepted a proposal, EPA and the developer work with the regional offices and state
agencies to identify a site which contains wastes suitable for testing the capabilities of the technology. EPA
prepares a detailed sampling and analysis plan designed to thoroughly evaluate the technology and to ensure that
the resulting data are reliable. The duration of a demonstration varies from a few days to several months,
depending on the length of time and quantity of waste needed to assess the technology.
Sometimes EPA conducts a series of tests to assess the performance of a specific technology on a specific
waste. This is called a treatability study. This was the mechanism used to test the McColl waste at the OES
research facility.
McCOLL SITE DESCRIPTION
The McColl Site in Fullerton, California, was used from 1942 to 1946 for the disposal of waste consisting
of acidic sludges from the alkylation and product-treating processes used in the refining of aviation gasoline. The
waste was deposited in twelve large pits (sumps). At that time, the property surrounding the site was relatively
undeveloped except for an oil field to the north, a hog farm, and agricultural land to the south. From the
mid-1950s to 1962 four of the six eastern sumps were covered with drilling mud from nearby oil fields. The
other six sumps were covered in the late 1950s by the construction of the Los Coyotes golf course. The site
currently consists of two distinct areas referred to as the Ramparts and Los Coyotes areas, as shown in Figure
2. In 1968, homes were built in the area approaching the eastern border of the Ramparts section of the site.
From 1978 to 1980, residential development was completed in the areas east, south, and north of the site. In the
late 1970s, local, state, and federal agencies investigated complaints from residents near ihe site about odors and
health problems believed to be related to the McColl Site. As an interim measure, the Ramparts area of the site
was covered in 1983, and odors from the site have since decreased. From 1980 to 1983, the California
-------
RALPH B.CLARK
REGIONAL PARK
| ou^ro ^_j
'
>,; LOS COYOTES
GOLF COURSE
FULLERTON
ROSECRAN
Figure 2. McColl Site layout.
8
-------
Department of Health Services (DHS) conducted a remedial investigation/feasibility study (RI/FS) to determine
the best course of action to be taken to remediate the site.
In 1984, the EPA and DHS decided that the appropriate action was excavation, removal, and
transportation of the waste to an off-site facility for final disposal. Execution of the remedy was halted in 1985
by a State court injunction requiring the preparation of an environmental impact report (EIR) prior to removal
of material from the site. In 1989, EPA completed an SROA which concluded that thermal destruction was a
viable method for the site remediation.
The three major waste types encountered at the McColl Site are 1) viscous, black, tar-like wastes; 2) grey
sludge-like material that may be drilling mud; and 3) hard, black, asphaltic wastes. The predominant type of
waste is the hard black asphaltic type. It is characterized by having low pH (2), high sulfur content (8.1%), and
elevated concentrations of organic sulfur, aromatics, and aliphatic hydrocarbons. The acid sludge has changed
in physical and chemical character over the years: the acid component in the sludge has reacted with the oil
component, producing a mixture of complex organic and sulfur-containing compounds. Some components of the
sludge have hardened or polymerized, causing a physical change in the waste material.
Borings were conducted in 1987 to characterize the McColl waste. Each boring was sampled and
analyzed for all of the organic and inorganic constituents on the hazardous substance list (HSL) excluding
pesticides. (EPA had researched the waste disposed at the site and determined that pesticides never were present.)
Table 1 lists the average and maximum concentrations of the organic and inorganic compounds identified in the
site waste. Based on these results EPA selected drums of waste to be used in the treatability study. Further
analyses of the candidate drums provided a final selection that was representative of the site waste and consistent
with the OES research facility's permit conditions. The site waste material was screened, divided into two sets
of feedstock (one blended with sand and the other left as raw waste), and transported to the OES facility for
processing.
PROGRAM OBJECTIVE
The objective of the demonstration was to evaluate the effectiveness of the CBC technology in thermal
destruction of the organic hazardous constituents from the McColl waste and thus to determine if the technology
is a viable candidate for further on-site testing under the SITE Program. During the test, OES operated and
maintained the CBC, and provided process monitoring and continuous emissions monitoring (CEM). EPA
conducted process inlet and outlet stream sampling and analysis to determine the CBC's operating efficiency and
contaminant destruction capabilities. Stack sampling and analysis were designed to provide the data necessary
to evaluate the CBC and assess the environmental effects of a full-scale SITE Demonstration Test using the CBC
technology. Ash analyses were designed to permit an evaluation of the disposal options; for the ash product.
The treatability study's scope was limited by the OES facility permits. An extended project scope,
development of more in-depth criteria, and collection of more extensive data will be required in any future SITE
Demonstration Test.
EPA will be evaluating the results of the demonstration to determine if the emissions levels can meet
SCAQMD requirements. If so, an on-site, full-scale test of the OES CBC technology would gather data about
this type of thermal treatment and determine its applicability, not only to the McColl Site, but also to other
Superfund site cleanups.
-------
TABLE 1. CHEMICAL CHARACTERIZATION OF MCCOLL WASTE
Constituent
Organics
Methylene Chloride
Acetone
Chloroform
2-Butanone
Benzene
2-Hexanone
Toluene
Ethyl Benzene
Total Xylenes
2-Methyl Phenol
4-Methyl Phenol
Benzoic Acid
Naphthalene
2-Methyl Naphthalene
Diethyl Phthalate
Flourene
Phenanthrene
Di-n-Butyl Phthalate
Bis(2-ethylhexyl)Phthalate
Thiophene
Others (b)
Metals
Arsenic
Barium
Beryllium
Chromium
Cobalt
Lead
Manganese
Nickel
Tin
Zinc
Average(a)
value
(ppm)
3
8
2
0.4
52
0.03
71
27
74
0.07
0.6
2
55
20
2
2
3
0.03
0.4
89
553
17
71
1
28
4
5
104
13
2
28
Maximum(a)
value
(ppm)
17
58
2
3
415
0.1
728
208
744
0.7
1
2
120
152
9
2
7
0.05
5
678
10664
203
517
1
171
8
15
242
75
7
43
(a) See Appendix V hi Field Notes for details.
(b) Organic compounds not identified on the hazardous substances list.
10
-------
PERFORMANCE CRITERIA
The criteria noted below were used to evaluate system performance and effectiveness, and to satisfy the
program objective.
Destruction and Removal Efficiency (DRE)
DRE was the basis for evaluating the efficiency of the combustion process in destroying organic
constituents while treating the McColl Site waste. The calculations for DRE were based on the destruction of
a performance indicator (CC14) added to the feed material and analyzed in the residue streams. CC14 was selected
because it is a difficult to destroy chlorinated compound that has been used as a performance indicator on previous
RCRA trial burns.
Flue Gas Emissions
Flue gas analyses are among the most important criteria for a thermal destruction process because they
evaluate the quality of the gas being emitted to the air. The treatability study was designed to evaluate acid gas
emissions to indicate the chloride content of the flue gas and to determine the capability of the system to capture
chlorides; particulate concentrations to show the amounts of solid material passing through the air pollution
control devices; flue gas organic concentrations to determine the degree of organics destruction, the amounts of
PICs and the in-process formation of new undesired hazardous organic components; inetals concentrations to
determine the disposition of volatile and particulate metals; and CEM results to quantify the criteria pollutants
(CO, THC, SO2, NOJ, oxygen (O^, and carbon dioxide (COJ present in the gas being emitted to the
atmosphere.
Organics Destruction
Organics (including those volatiles listed on the HSL), semivolatiles, dioxins, furans, and PCBs were
analyzed in the feed, bed ash, and fly ash
streams, and in the stack gases. These analyses indicated the quantity of organic substances remaining in the
residual streams, and thus, whether or not the process residuals had been rendered non-hazardous.
Toxic Metals Distribution
The feed, residue, and flue gas streams were evaluated for the presence of toxic metals to determine 1)
the partitioning of the non-volatile metals into the resulting ash; 2) stabilization within the ash matrix to prevent
the metals from leaching; and 3) capture of the volatile metals from the flue gas stream before their release into
the air.
TCLP Results
TCLP tests measured metals and organics in the ash to determine if it could be designated as
non-hazardous.
Physical Parameters
Several physical parameters of the waste feed and ash streams were evaluated to characterize the fly ash
and bed ash residues.
11
-------
System Operating Conditions
The process operating parameters were tracked and logged to allow 1) process calculations; 2) evaluation
of system problems during operation; 3) determination of equipment maintenance requirements; 4) quantification
of waste residuals; and 5) development of costs to process the material.
OA/OC
Laboratory quality control samples were collected to evaluate if the precision, accuracy, and completeness
objectives of the QAPP had been achieved.
KEY CONTACTS
For more information concerning the treatability study, the following persons may be contacted:
USEPA Project Manager
Mr. Douglas W. Grosse
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
513-569-7844
Region DC Contact
Mr. John Blevins
Ms.' Pamela Wieman
U.S. Environmental Protection Agency
Region IX
75 Hawthorne St.
San Francisco, CA 94105
415-744-2241
OES Test Coordinator
Mr. Jeffery Dasch
Ogden Environmental Services, Inc.
Director of Operations and Technical Support
3550 General Atomics Court
San Diego, CA 92121
619-455-3045
12
-------
SECTION 3
VENDOR'S SYSTEM DESCRIPTION
INTRODUCTION
The information in this section is a presentation provided, reviewed, and approved by Ogden
Environmental Services, Inc. The EPA makes no claims as to the accuracy of the statements.
SYSTEM DESCRIPTION
The nominal 2 MM Btu/hr CBC used in the SITE treatability study is owned! and operated by Ogden
Environmental Services, Inc., San Diego, California.
The primary CBC components are the combustion chamber, hot cyclone collector and solids return
system, flue gas cooler, baghouse, and stack. Auxiliary systems include feed material systems, auxiliary fuel
system, forced-draft and induced-draft fans, bed ash removal conveyor, compressed air system, cooling tower,
safety and process control equipment, off-gas analysis systems, and building ventilation. Figure 3 is a schematic
presentation of the process.
The OES pilot-scale combustor is 27 ft high with a 16-in. inside diameter and 13-in. thick refractory
liner. The CBC loop components (combustor, cyclone, and loop seal) are constructed of refractory-lined carbon
steel which provides a thermal barrier as well as corrosion and erosion resistance.
Air System
A positive displacement blower pushes primary combustion air into the CBC below the air distributor and
injects secondary combustion air through various ports along the combustor wall. The primary and secondary
air flows are controlled independently. The primary air is pumped into the lower portion of the combustion
chamber where the bed material is fluidized by turbulent mixing of the air and solids. Larger waste particles
gravitate downward to form a more dense fluidized bed in the lower combustor zone. Smaller waste solids are
carried up to the top of the combustor. The secondary air is supplied to various locations in the combustion
chamber to ensure complete combustion of the waste and to minimize formation of nitrogen oxides (NO*).
The high velocity of the combustion air and circulating waste create a uniform temperature (± 50°F)
around the combustion loop, which is controlled at a value between 1,425°F and 1,800°F. This results in
efficient combustion and eliminates the need for an afterburner. Residence times in the combustor range from
approximately 1.5 sec for gases to 30 min for waste material. The combustion and neutralization of acid gases
within the combustor chamber eliminate the need for high-alloy combustor components or post-burner treatment
units - such as wet scrubbers — and effect the desired contaminant destruction and acid- gas capture. CO and
NOX emissions are controlled to low levels by 1) the mixing of the fuel, waste feed, and combustion air; 2) the
combustor's relatively low operating temperatures; and 3) staged secondary air injection along the combustion
chamber length at progressively higher locations, resulting in continued combustion and destruction of waste over
the entire height of the combustion chamber.
The high turbulence of solids within the combustor makes the CBC combustion process relatively
insensitive to feed properties. Wet waste feeds and slurries can be fed and incinerated as easily as dry solids.
The high internal combustor heat transfer ensures that moisture in the feed evaporate!? with little depression in
local combustor temperature. One effect of feed moisture on the incineration process is on the heat energy
balance: energy that would otherwise go to processing contaminated soil is used to evaporate water. The CBC
13
-------
COMBUSTOR
STACK
(Courtesy of Ogden Environmental Services, Inc.)
Figure 3. Schematic of the OES CBC system
is insensitive to large amounts of fines in the feed streams — feed fines benefit the waste circulation that produces
the isothermal combustor conditions. The combustor's design controls the behavior of fines automatically.
Fuel System
An auxiliary fuel system — capable of up to 2 MM Btu/hr thermal output — supplies natural gas to the
ignitor during start-up and to the in-bed lances during steady-state operation. The natural gas provides process
heat during waste treatment. An operator adjusts the natural gas flow rate to control the flue gas oxygen
emissions at their target value. Retractable, water-cooled, bayonet heat exchangers are available, when needed,
to remove heat from the combustor.
Solids System
A fully-enclosed solids feed station is used to prepare the waste feed. The station includes a ventilated
alcove where drums are unsealed and waste is typically sampled before it is loaded into the feed bunker. The
station is maintained at negative pressure to minimize fugitive emissions. Ventilation gases are pulled through
activated carbon and HEPA filters before they are exhausted.
Waste is transported pneumatically from drums to a solids surge bunker and then is dropped into the feed
bunker. A weigh cell continuously displays the inventory remaining in the feed bunker. (The San Diego APCD
allowed a 200 Ib/hr maximum feed rate for the treatability study.) An Acrison screw-auger meters the waste from
the feed bunker to a rotary valve and into the CBC loop at the solids return leg where it contacts the hot waste
stream recirculated by the cyclone. Key operating parameters are monitored continuously. If the
permit-mandated limit is reached, the feed system stops automatically.
14
-------
Auxiliary fuel and waste feed are introduced separately into the lower combustion chamber. Dry
limestone sorbent is added to control gaseous emissions of sulfur and chloride. Sorbent feed is loaded manually
into a bunker from which it is metered by a screw-auger to a rotary valve and into the CBC. A level sensor
measures the inventory of sorbent in the bunker. The sorbent feed rate is adjusted manually to control acid-gas
emissions below the permit- mandated limit. Elutriated solids are separated from the flue gas by a hot cyclone
and reinjected into the lower combustor by means of a proprietary non- mechanical seal. The continuous ignition,
burning, and reaction of the fuel, waste feed, sorbent, and ash components effectively destroy the waste's contam-
inants. Acid gases and sulfur oxides formed during incineration are captured by the limestone sorbent, which
combines with hydrochloric acid to form calcium chloride, and with SOX to form calcium sulfates.
During operation, bed ash is removed periodically from the CBC by means of a water-cooled ash removal
system consisting of a screw conveyor and bucket elevator. The bed ash is cooled to below 350 °F to allow safe
discharge into drums. Negative pressure ventilation prevents the release of fugitive bed ash dust.
Flue Gas System
The temperature of the hot gas leaving the cyclone is lowered below 400 °F in a water-cooled flue gas
cooler (FGC). The FGC heat transfer surfaces are cleaned on-line by soot blowers. A high-pressure blower,
fixed position nozzles, and automatic sequencing controls and valves are provided for tiiie soot-blowing system.
The fly ash that is carried through the FGC is collected continuously by baghouse filters. These filters
reduce the particulate loading in the flue gas below the EPA RCRA particulate emission limit of 0.08 gr/scf. The
system consists of two baghouse modules, each of which contains 25 fabric filters with 275 sq ft of filter surface
area. The filters, cleaned on-line by pulse jets, are acid-resistant and can operate at temperatures up to SOOT.
If necessary, either baghouse module can be removed from service without interrupting testing or compromising
safety. The baghouse features are described more fully in Table 2.
An induced-draft (ID) fan is located directly upstream from the stack. The ID fan suction is adjusted
automatically. It maintains the entire combustion loop, the flue gas cooling system, and the cleaning equipment
slightly below atmospheric pressure to prevent fugitive emissions.
Exhaust Air System
The research facility is serviced by a dedicated ventilation system providing 7,500 cfm of air flow. All
exhaust air is HEPA-filtered to control the release of fugitive dust.
PROCESS CONTROL
System control functions are exercised from a central control room. Microprocessor-based proportional
integral derivative (PID) controllers govern process flows such as system pressures and combustion air, auxiliary
fuel, and waste feeds. Thermocouples monitor combustor temperatures through thermowells penetrating the
combustor shell and refractory. A control panel provides remote readout of the combustion temperatures.
Combustion air is injected below the air distributor and at various levels along the chamber wall. Primary
air is maintained automatically at a specified flow rate. Secondary air is also controlled! automatically — with a
butterfly valve on the main secondary air supply header and by individual throttling valves in each secondary air
injection pipe. Air flow rates are measured by annular flow meters. The ratio of secondary to primary air can
be controlled manually, as indicated by the particular test requirements and by flue gas emissions. The ID fan
is adjusted automatically to achieve the desired pressure balance at the waste feed inlet. The control panel gives
remote readouts from a pressure sensor at the feed point and from differential pressure sensors at the gas
distributor, the riser section of the combustor, the hot cyclone, and the non-mechanical seal.
15
-------
TABLE 2. BAGHOUSE FILTER SYSTEM
Operating Conditions
Temperature <500°F maximum
Gas volume 800 acfhi
Dust loading leaving filter <0.08 gr/dscf
Dust removal efficiency 99.6%
Maximum pressure drop 15" we (baghouse cleaning will keep DP <6" we)
Design Criteria
Type of bag cleaning Pulse jet/on-line (off-line as backup)
Gas flow pattern Draw through/suction
Maximum air to cloth ratio 3.2 acfm/sq ft
Bag type Outside collecting
Bag material Gore-Tex membrane filter bags
(total filtering surface 550 sq ft)
Cage material Carbon steel
Number of modules 2
A speed-controlled metering screw conveys waste to the combustor feed point where the waste contacts
hot circulating solids. This mixed waste is then injected into the combustion chamber. The natural gas fuel
system used for start-up and for auxiliary heat is normally off during waste burning. If the waste feed is
interrupted, the auxiliary fuel system will be reactivated. This feature allows combustion temperature control
during brief, intermittent, and correctable fuel feed interruptions.
Safety Controls
The feed rate controllers are interlocked to key process parameters through the programmable logic
controller (PLC). These key process parameters are identified in Table 3. If any of the measured variables
exceed permit-mandated or acceptable limits, or in the event of a power failure, the feed stops automatically.
If the emergency stop button is pressed, all feeds and fluidizing air are shut off, while the ID fan is left
on to maintain negative pressure on the system, thus minimizing fugitive emissions. Auxiliary fuel is controlled
by a valve train that is interlocked to bed temperature, start-up burner flame-detection, flue gas oxygen,
combustor pressure, cooling-water flow, and total air flow. If any of the measured parameters exceed acceptable
limits, the valve train will shut off the auxiliary fuel. Valves with fail-safe positions are designed to enclose all
solids and vapors within the combustor system.
The critical system equipment is monitored automatically, and conditions are displayed on the control
panel, which is staffed by trained operators. Equipment failures, which could compromise system performance
and safe operation, either initiate automatic control/corrective procedures, or alert the operator to the system
problem through an alarm annunciator. This, coupled with frequent routine inspections, serves to adequately
mitigate the effect of any equipment failure.
16
-------
TABLE 3. INTERLOCK SETTINGS
Parameter
Set Point
CO concentration (5-minute average)
Combustion temperature (1-minute average) .
Combustion residence time (1-minute average)
O2 concentration (5-minute average)
THC concentration (5-minute average) . . . .
SO2 concentration
Baghouse inlet temperature
1,000 ppm (volume)
<1,300°F
< 1.0 sec
... 1.0 volume %
400 ppm (volume)
504 ppm (volume)
>400°F
PROCESS MONITORING AND DATA ACQUISITION
All significant process data are monitored and logged either on manual logs, six 3-pen strip recorders,
a 30-channel chart recorder, or the 256-channel computerized data acquisition system (DAS).
The computerized DAS displays key process variables on CRTs located in the control room. These
variables include loop temperatures and flue gas composition. The system prints out selected data on a line
printer at 15 min intervals. The data include combustion air flows; natural gas flow; concentrations of O2, NOX
(as NO), SO2, CO, CO2, and THC; baghouse and combustor differential pressures; and combustor loop and inlet
baghouse temperatures. The DAS also can be used to analyze and plot data after a test's completion. Graphic
and tabular summaries of the treatability study data collected in the DAS have been prepared by OES and are
included as Appendix III in the field notes.
One chart recorder has 30 plotting channels. These process variables are described in Table 6 of
Appendix III in the field notes. Additionally, five 3-pen strip chart recorders continuously retain 1) bed
temperatures, O2 probe readings, and THC analyzer outputs; 2) CO, NOX, and CO2 analyzer outputs; 3) SO2 and
O2 analyzer outputs; 4) primary air, secondary air, and reactor purge data; and 5) cyclone pressure and natural
gas flows. Strip chart recordings are maintained in OES's files. Waste feed, limestoine feed, fly ash, and bed
ash logs are maintained manually to determine process rates.
17
-------
-------
SECTION 4
TEST PROGRAM PROCEDURES
WASTE FEED PREPARATION
In December 1988 selected drums of McColl waste were sampled to characterize the drum contents and
to select candidate drums that were both representative of the McColl Site waste and within Ogden Environmental
Services (OES) permit restrictions. Each drum was analyzed for total sulfur and total organic carbon (TOC); a
composite sample was further analyzed for volatile and semivolatile organics, dioxins, farans, pesticides, PCBs,
and metals. Results of the analyses are included in Appendix IV of the field notes.
After selection of the candidate drums, waste feed underwent preparation on-site in early March 1989 to
allow processing with a minimum of handling at the OES facility. The waste feed was inspected visually to
ensure that it was dry and free-flowing, then screened of debris, blended with sand if the sulfur content was in
excess of 5% (to meet permit restrictions), placed into clean DOT 17-H 55-gallon drums, sealed, labeled,
manifested, and shipped to the OES research facility on March 9, 1989.
OPERATING CONDITIONS AND OPERATING RANGE
The treatability study was conducted under processing conditions as close as possible to those expected
in a full-scale CBC. Process conditions were adjusted to control stack emission concentrations of criteria air
pollutants (CO, NOX, SO2, THC) and particulates while optimizing process rates for air and limestone. However,
regulatory restrictions affected system optimization.
The CBC was heated to an operating temperature of 1,700°F with an inert starting bed of sand and held
at this temperature over 20 hours to establish thermal equilibrium within the combustor prior to introducing the
waste feed. An 8-1/4 hr fine-tuning pretest to establish steady-state operating conditions was followed by Test
1, an 8-1/2 hr steady-state operation using blended, unspiked feed; Test 2, a 7-1/4 hr operation using unblended,
unspiked feed; and Test 3, a 6-1/2 hr operation using unblended, spiked feed. The a.ctual test durations are
shown in Table 4; process conditions are listed in Table 5. An after-test operation using blended, unspiked waste
was planned to gather additional CEM data but was not executed because the SO2 analyzer was not operating
within performance specifications.
Waste feed and limestone rates were controlled to ensure that emission concentrations of criteria pollutants
were maintained within permitted levels. Several SO2 spikes were experienced, activating the interlocks and
stopping the feed. At least four SO2 excursions occurred causing further feed interruptions. By late afternoon
it became apparent that the waste feed properties required a larger auger to attain the desired feed rates. The
auger was replaced during the night.
Test 1 began on March 29. Up to eleven SO2 spikes occurred, causing wastes feed interruptions. A
hydrocarbon spike also caused a feed interruption, which delayed sampling for about one hour. Tests 2 and 3
were conducted on March 30. OES decided to saturate the bed with limestone to prevent any further SOL, spikes.
Test 3 was conducted in the afternoon using unblended McColl waste spiked with 6,000 ppm (by wt.) of CC14.
Each day's test began with limestone feed, followed by waste feed. Operation during the first test
attempted to optimize the limestone-to-feed ratios, but SO2 spikes resulted. In the interest of staying well within
the permit limits, OES decided to use an excess of limestone. Combustion air and secondary air flows were
adjusted to control CO, CO2, and NOX emissions. Process conditions were maintained constant throughout the
sampling activity to ensure that adequate sample volumes were collected to obtain reliable analytical results.
Except for several momentary feed cutoffs, the system remained operational throughout the sampling runs. At
19
-------
TABLE 4. TEST DURATIONS
Day
3/27/89
3/28/89
3/29/89
3/30/89
3/30/89
Test
System
heatup
Pretest
fine tuning
#1
#2
#3
Duration
(hrs)
21-1/2
8-1/4
8-1/2
7-1/4
6-1/2
Feed
Bed ash/
sand
Blended
with sand
Blended
with sand
Unblended
Unblended
feeedrate
Spiking (Ib/hr)
None
Unspiked 400
Unspiked 325
Unspiked 170
Spiked 197
(6000 ppm
CC14)
Waste
rate
(Ib/hr)
None
200
163
170
197
Sampling
CEM only
CEM only
4 hours
full
sampling
3 hours
full
sampling
3 1/2
hours full
sampling
the conclusion of each day's test, the waste feed was stopped and the system was idled at 1700 °F until the next
day's operations. Additional daily events logged by ATC and OES can be found in Appendices II and III of the
field notes.
SAMPLING AND ANALYSIS PROGRAM
ATC was responsible for collection and analysis of all program samples including QA/QC samples. OES
was responsible for ensuring uninterrupted operation of the CBC system, collection of all CEM data, monitoring
of process parameters, and maintaining weight logs for waste feed, limestone feed, bed ash, and fly ash.
The CBC process stream monitoring locations are shown in Figure 4. Table 6 describes the OES
monitoring parameters and methods, and Table 7 summarizes the ATC process stream sampling and analysis
parameters and methods. Changes to the planned ATC program are summarized in Appendix II of the field
notes.
Process Stream Sampling
Waste Feed ~
Grab samples of the waste feed were taken during the flue gas sampling tests, composited, and riffled
prior to aliquoting for the different analyses. Discrete grab samples were taken for volatile analysis on an hourly
basis, yielding a total of eight discrete samples for the test. The samples were collected at a point after the feed
hopper, just before the waste material entered the combustion chamber.
20
-------
TABLE 5. TREATABILITY STUDY PROCESS/PERMIT CONDITIONS
Process
conditions
Waste feed rate
Limestone feed rate
30 mesh sand
Fly ash rate
Bed ash rate
Carbon tetrachloride
Mid combustor temp.
Residence time
Superficial velocity
Combustion air
Natural gas
Flue gas oxygen dry
Flue gas oxygen wet
CO emissions(d)
NOX emissions
S02 emissions
HCl emissions
THC emissions(d)
Particulates5.5%
ns
<100 ppm
100 ppm
<500 ppm
0.2 kg/hr
<100 ppm
<0.08 gr/dscf
<6 in wg
300°-375°F
ns
ns
ns
Federal/State
Permit
limits
<500 lb/hr(c)
ns
ns
ns
ns
ns
>1425°F
>1.4 sec
ns
ns
ns
>5.3%
ns
<250 ppm
ns
ns
<1.8 kg/hr(g)
<100 ppm
<0.08 gr/dscf
<12 in wg
<400°F
ns
ns
ns
APCD
Permit
limits
200
ns
ns
ns
ns
ns
>1425°F
>1.4
ns
ns
ns
>5.3
ns
<250
ns
<504(e)
ns
<100 ppm
ns
<15 in wg
<400°F
ns
ns
ns
Test 1(a)
325(b)
197
162
231
201
0
1721
1.54
17.5
331
19.6
11
9.1
30
49
(f)
<0.0040
5
0.0029
3.1
377
979
868
923
Test 2(a)
171
349
0
296
53
0
1726
1.52
17.7
331
23.1
9.9
7.9
30
58
(f)
<0.0035
1
0.0035
3.5
368
957
850
903
Test 3(a)
197
182
0
201
21
0.22
1709
1.55
17.4
331
18.7
11.8
9.7
26
48
(f)
<0.0037
2
0.0023
4.9
357
984
905
946
(a) Average values for comparison only. NOT TO BE USED FOR CALCULATIONS. See Appendix II and III in Field Notes
for details.
(b) Includes 50% sand. Permit restriction of 200 Ib/hr refers to waste only.
(c) Feed was restricted to 5% total sulfur content.
(d) Dry basis.
(e) APCD restricts flue gas S02 concentrations (corrected to 3% 02 dry standard conditions) to less than the
equivalent S02 concentration when the fuel composition is corrected to 0.5% sulfur by weight.
(f) Not quantifiable. S02 analyzer was not operating within performance specifications.
(g) . Or 99% removal efficiency.
(h) Corrected to 7% 02.
ns not specified
21
-------
TABLE 6. DESCRIPTION OF CBC MONITORING PARAMETERS AND METHODS
Process parameter
Combustion air flow rates
Natural gas flow rates
Limestone feed rate
Solid waste feed rate
Bed ash drain rate
Fly ash drain rate
Combustion loop temperatures
Combustion loop pressures
Combustion loop DPs
Flue gas outlet temperature
Flue gas oxygen
Baghouse DP
Extractive flue gas composition
02
CO
C02
NO.
S02
HC
Location""
1A, 1B, 1C
2
3
4
5
6
7
8
8A
9
10
11
12
Instrument type
Pilot tube plus DP cell
Mass flow meter
Level detector
Loss- of -weight-bunker
Barrel scale
Barrel scale
Type K thermocouples
DP cells
DP eel Is/DP Gauges
Type K thermocouples
02 probe
DP cell
Paramagnetic
Infrared
Infrared
Chemical luminescence
Infrared
Flame ionization
detector
Recording
frequency
Recorder"" (per hr)
YEW, DAS, CR
YEW, DAS, CR
Vibrascrew log 1
DAS, Acrison log 2
Bed ash log 1
Fly ash log 1
YEW, DAS, CR
YEW, DAS, CR
YEW, DAS, CR
YEW, DAS,
YEW, DAS, CR
YEW, DAS,
YEW, DAS, CR
YEW, DAS, CR
YEW, DAS, CR
YEW, DAS, CR
YEW, DAS, CR
YEW, DAS, CR
(a)
See Fig. 4
(b) Recording media: YEW = 30 channel recorder (every 8 s); DAS = computerized data acquisition system (every 2
min); CR = chart recorder (continuous); logs = manual data.
22
-------
TABLE 7. PROCESS STREAM SAMPLING AND ANALYSIS
Locating and
sampling
parameter Location(a)
Waste feed 4
Bed ash and 5
Fly ash 6
(2 streams)
Limestone feed 3
(a) See Figure 4.
Analytical
parameter
Metals (b)
Vol. w/lib scan
Carbon tetrachloride
SV/PCBs w/scan
PCDD/PCDF (c)
Moisture
Ash
Ultimate analysis
Heating value
Density
Sulfur, total
Total organic carbon
pH
Organic halogens
Inorganic halogens
Metals (b)
Vol. w/lib scan
Carbon tetrachloride
SV/PCBs w/lib scan
PCDD/PCDF (c)
Moisture
Ash
Density
Sulfur, total
pH
Organic halogens
Inorganic halogens
TCLP vo I at iles
TCLP metals, Cr(VI)(d)
Inorganic halogens
Organic halogens
Metals (b)
Calcium
(b) Total metal parameters include: Sb, As, Ba, Be, Cd, Cr, Co, Cu, Pb, Mn,
and Zn.
(c) Dioxin/furan groups include tetra- through
(d) TCLP metals include: As, Ba, Cd, Cr, Pb,
octa- groups and confirmation
Hg, Se, Ag.
Analytical
methods
EPA 3050/3010/7000
EPA 8240
EPA 8240
EPA 8270/680
EPA 8290/ASME
D 95
D 482
D 3178, E 258
D 240
D 1298
E 395
Leco
EPA 9040
EPA 8240, 300.0
ASTM E776, EPA 300.0
EPA 3050/6010/700
EPA 8240
EPA 8240
EPA 8270/680
EPA 82:90/ASME
D 1744
D 482
D 1298
E 395
EPA 9040
EPA 8240, 300.0
ASTM E776, EPA 300.0
SW 846 8240
SW 846 6010/7000/7196
Extraction, EPA 9020
Leacheite, EPA 9020
EPA 3050/3010/7000
Hg, Hi, Se, Tl, Ag, Sn,
of 2, 3, 7, 8-TCDD and TCDF.
23
-------
SOUD _
WASTE
AUXILIARY FUEL
COMBUSTION AIR
(Courtesy of Ogden Environmental Services, San Diego, Calif.)
Figure 4. Sampling and monitoring points of the CBC system.
Limestone Feed —
A grab sample from the limestone feed chamber was taken at the beginning and end of each test, designed
to yield three discrete samples. However, due to a field oversight, limestone samples were not taken during Test
2.
Combustor Bed Ash —
Bed ash material was discharged from the bottom of the CBC and deposited into 55-gal drums; a
time-integrated grab sample was taken from each drum, composited, and riffled. Volatile samples were collected
as separate discreet samples and composited.
Fly Ash -
Fly ash was discharged from the baghouse into 55-gal drums. Core samples were taken from each drum,
then composited, riffled, and aliquoted. Discreet samples were collected and composited for the volatile organics
analyses.
Initial collection both of combustor ash and fly ash was attempted with a thief sampler. Because the
sampler was unable to penetrate the drum depth without causing binding, the sampling procedure was revised for
the first test by vacuuming the collection drum's contents into another drum and taking grab samples at equal
depths throughout the drum. During Tests 2 and 3, 30-min grab samples were collected from the 55-gal drums.
24
-------
TABLE 8. FLUE GAS SAMPLING AND ANALYSIS
Sampling
parameter
Analytical
parameter
Analytical
parameter
Paniculate
Semivolatile Organics
Volatile Organics
Chloride
CO2/O2/CO/NOX
SO/THC
Participate
Metals(a)
SV/PCB
PCDD/PCDF (b)
Vol. w/library scan (c)
Carbon tetrachloride
Chloride
CO2/O2/CO/NOX
SO2/THC
EPA M5 Gravimetric
EPA 3050/3010/7000
EPA 8270/680
EPA 8290/ASME
EPA 5040
EPA 5040
EPA 300.0
CEM System
CEM System
(a) Metal parameters include: Sb, As, Ba, Be, Cd, Cr, Co, Cu, Pb, Mn, Hg, Ni, Se, Tl, Ag, Sn,
and Zn.
(b) Dioxin/furan groups includetetra- through octa- groups and confirmation of 2, 3, 7, 8-TCDD and TCDF.
(c) Four VOST pairs were collected per 4-hr run. Three pairs were analyzed and one pair held as backup.
Analysis of VOST condensate is contingent upon flue gas moisture content.
Stack Sampling
The OES stack was modified to permit the use of several sampling trains to collect flue gas samples. A
Modified Method 5 (MM5) train was used to collect particulates, metals, and semivolatile samples; a volatile
organic sampling train (VOST) was used for organics; and a separate gaseous hydrogen chloride train was used
for HC1. Gas rates were determined using a pitot tube. Table 8 summarizes the flue gas sampling and analytical
parameters and methods.
ATC sampling and analysis activities determined the properties of the waste feed, limestone, bed ash, fly
ash, and flue gas streams. The analytical workload was divided among three laboratories: Clean Harbors
Analytical Services, Bedford, Mass, (formerly the Alliance Technologies Chemistry Division); Triangle
Laboratories, Research Triangle Park, N.C.; and Industrial Testing Laboratories, St. Louis, Mo. Industrial
Testing Laboratories was responsible for analysis of various physical parameters such as density, heating value,
moisture, viscosity, ultimate analysis, pH, TOC, sulfur, and ash; Triangle Laboratories analyzed the flue gas
samples for dioxins, furans, semivolatiles, and PCBs; and Clean Harbors Analytical Services was responsible for
all other analyses, including volatiles, trace metals, total chlorides, HC1, organic and inorganic halogens, sulfur,
TCLP results, calcium, particulates, and CC14.
The customized Method 5 train sampling was conducted isokinetically from 12 points in the stack for 15
min per point. Sampling was conducted while traversing the train across a length equal to two stack diameters.
Flue gas parameters were recorded every 10 minutes. The VOST was operated at a flow rate of one L/min for
20 min. Four 20-min samples constituted one VOST run. Three runs were completed to yield 12 sets of VOST
25
-------
tubes, not including QA/QC sets.
The gaseous hydrogen chloride sampling train was operated for each sampling run, and semivolatile
organic compounds including PCBs, PCDDs, and PCDFs were collected in an MM5 train. The train was
operated for 20 min per point across a 12-point traverse.
Each train was leak-checked before and after each sampling run, and leakage rates were documented on
the relevant field test data sheet. Recovery procedures for each train are documented in the Treatability Study Test
Plan [61.
The Continuous Emissions Monitoring system (CEM) was maintained and operated by OES. The CEM
sampling ports were located directly downstream of the baghouse filter modules. Particulates and moisture were
removed from the flue gas sample before the sample was pumped into the continuous on-line analyzers. The flue
gas analysis system provided an accurate near-real-time indication of flue gas composition. Provisions were made
for on-line and off-line calibration of analyzers. The flue gas analysis system consisted of an extractive analysis
system for monitoring O2, CO2, CO, NOX, SO2, and THC concentrations, and an in situ system for monitoring
flue gas oxygen concentrations. The hi situ oxygen probe was used to measure oxygen concentration at the exit
of the flue gas cooler. The flue gas analyzers used to measure the flue gas composition are identified in Table
9.
EPA sampling methods were used to collect flue gas and particulate samples specific to each test and test
feed material. Prior to planned operation of the research facility CBC, OES quality assurance performed a review
and inspection of each of the CEM system transmitters and recorders system, as well as other essential control
instruments to ensure that they were within designated accuracies. In addition, the EPA arranged for an
independent audit of the CO, NOX, and SO2 analyzers as part of the test program. The operation of the analyzers
is discussed hi Section 5, and audit results are presented in Appendix II of the field notes.
Analytical Procedures
Physical/Chemical Analysis —
Waste feed samples were analyzed at Industrial Testing Laboratories for moisture, ash, heating value,
ultimate analysis, density, total sulfur, TOC, and pH.
Inorganic Analysis —
Metals — Metals analyses were prepared using SW-846 methods 3010 or 3050. Occasionally certain
wastes were not exactly amenable to these procedures, resulting in the need for such procedural modifications
or alternative preparation procedures as 1) the use of a reduced sample size to provide elevated detection limits;
2) addition of small quantities of sulfuric acid in Method 3050 or hydrogen peroxide in Method 3010 to aid
dissolution; or 3) elimination of hydrochloric acid from both Method 3010 or 3050. The need for these modifi-
cations was determined after sample preparation, and is documented in Appendix II of the field notes.
Analyses for all metals except mercury and hexavalent chromium were performed using
inductively-coupled argon plasma emission spectroscopy (ICAP). Mercury analysis was performed using either
Method 7470 or 7471, and hexavalent chromium in TCLP extracts was determined colorimetrically. Detection
limits for these analyses are provided in the tables in Appendix II of the field notes.
Organic and inorganic halogen and hydrogen chloride — Soil and ash samples were analyzed for chloride,
bromide, and iodide by ion chromatography (1C). The preparation methods used various analytical techniques
in a consolidation designed for this program. The procedures have not been tested relative to the precision,
accuracy, detection limits, and anticipated method blank concentrations.
26
-------
TABLE 9. FLUE GAS ANALYZERS
Parameter
02
Co2
NOX
CO
SO2
THC
Analyzer type
Beckman analyzer, Model 755
Lira analyzer, Model 3000
Beckman analyzer, Model 951 A
Beckman analyzer, Model 865
Beckman analyzer, Model 865
Beckman analyzer, Model 400A
Range
0-25%
0-25%
0-1,000 ppm
0-1,000 ppm
0-1,000 ppm
0-1,000 ppm
TCLP metals — Ash samples were prepared using the extraction procedure detailed in the Federal
Register, Vol. 51, No. 216, November 7, 1986, pp. 40643-53. Analyses were performed for RCRA metals
including arsenic, barium, cadmium, chromium, chromium (VI), lead, mercury, selenium, and silver.
Organic Analysis —
Volatile organics - The waste feed, bed ash, fly ash, and flue gas samples were subjected to volatile
organic analyses for CC14 and the compounds listed in Table 10. A library search was conducted to determine
tentatively identified compounds (TICs) by comparing the unknown spectra to referenced spectra in the
EPA/NBS/NIH library.
Semivolatiles/PCDD/PCDF/PCBs - The semivolatile analytes included the HSL organics shown in Table
11. The PCDD/PCDF analyses provided concentrations of the tetra through octachlorinated homologues, and
confirmation of the 2,3,7,8-isomers of TCDD/TCDF. The PCB analyses provided determinations of the mono
through decachlorinated congeners.
In order to accommodate the analysis of SVs, PCBs, and PCDD/PCDFs from the same MM5 train, a
two-stage extraction of the filter and XAD resin was used. This multiple analysis/extraction scheme was a
combination of several approved EPA methods but the combination has not been validated. The data obtained
using this method was flagged and evaluated in terms of the technique's questionability.
Organic halogens and sulfur — Organic halogen and organic sulfur compounds were identified and
quantified from the volatile and semivolatile GC/MS analysis of test samples.
TCLP volatile organics — Ash samples were analyzed for TCLP volatile organics using the
zero-headspace extraction procedure detailed in the Federal Register under TCLP metals. Analysis of the filtrates
for volatile organics listed in Table 12 was also conducted.
QA/QC PLAN AND AUDITS
The QA/QC protocols followed for this program are outlined in the Treatability Study Test Plan [6].
Table 13 details the QC samples and blanks used to validate the data including method blanks, calibration check
samples, matrix spikes, matrix spike duplicates, laboratory control samples, and surrogate spikes. Table 14 lists
the frequency for each measure.
27
-------
TABLE 10. LIST OF VOLATILE ORGANICS
Bromodichloromethane
Benzene
Bromoform
Bromomethane
Carbon tetrachloride*
Chlorobenzene
Chloroethane
Chlorofonn
Chloromethane
cis-l,3-Dichloropropene
Dibromochloromethane
1,1-Dichloroethane
1,2-Dichloroethane
1,1-Dichloroethylene
trans-1,2-Dichloroethylene
trans-l,3-Dichloropropane
1,2-Dichloropropane
Ethylbenzene
Methylene chloride
1,1,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1,1,1 -Trichloroethane
1,1,2-Trichloroethane
Trichloroethylene
Trichloroflouromethane
Vinyl chloride
Performance indicator compound
The analytical data was validated by the individual laboratory's QC group using the criteria outlined in
the Test Plan. In addition, ATC used results from field and laboratory method blanks, replicate samples, and
internal QC samples to further validate the analytical results. They employed field blanks and replicate'field
samples to validate sample collection. The following criteria were used to evaluate field sampling data:
implementation of approved test procedures; proper operation of the process being tested; maintenance of properly
operating and calibrated equipment; performance of leak checks before and after tests; selection of reagents that
conformed to QC specified criteria; use of NBS traceable CEM calibration gases; adherence to proper chain-of-
custody; and verification of proper sample gas volume collection in VOST, M5, and MM5 trains.
Analytical data were evaluated using the following criteria: approved analytical procedures; properly
operating and calibrated instrumentation; acceptable results from analyses of QC samples (i.e., the reported values
had to fall within the 95% confidence interval for the samples); and achievement of precision and accuracy
comparable to previous analytical programs and consistent with the Test Plan objective.
^ To evaluate the reliability of the data, precision, accuracy, and completeness objectives were established.
Precision defines the degree of agreement among individual measurements made under prescribed conditions.
Accuracy compares a measurement to an accepted reference or true value. Completeness determines the percent
of samples judged to be valid.
The accuracy of this study was estimated through the analysis of laboratory control samples (LCS),
matrix-spiked samples, and surrogate-spiked components. Completeness was determined by comparing the
accuracy data to the recovery objectives in the Test Plan. Precision was verified through the analysis of duplicate
matrix spikes for selected matrices.
Corrective actions were taken at several points during the program. They fell into two categories:
immediate corrections that were implemented by the supervisor or analyst; and long term corrections that initiated
a formal paper trail. Immediate corrections in the field most often result from equipment failure or operation
oversight. They usually involve revaluation, reanalysis, or repeating a sample run. Analytical corrections
include recalibration of instruments, and reanalysis of QC samples or field samples.
28
-------
TABLE 11. HSL SEMIVOLATILE ORGANICS
Phenol
bis(2-Chloroethyl)ether
2-Chlorophenol
1,3-Dichlorobenzene
1,4-Dichlorobenzene
Benzyl alcohol
1,2-Dichlorobenzene
2-Methylphenol
bis(2-Chloroethyoxy)methane
4-Methylphenol
N-Nitroso-di-n-propylamine
Hexachloroethane
Nitrobenzene
Isophorone
2-Nitrophenol
2,4-Dimethylphenol
Benzole acid
bis(2-Chloroethyoxy)methane
2,4-Dichlorophenol
1,2,4-Trichlorobenzene
Naphthalene
4-Chloroaniline
Hexachlorobutadiene
4-Chloro-3-methylphenol
2-Methylnaphthalene
Hexachlorocyclopentadiene
2,4,6-Trichlorophenol
2,4,5-Trichlorophenol
2-Chloronaphthalene
2-Nitroaniline
Dimethylphthalate
Acenaphthylene
2,6-Dinitrotoluene
3-Nitroaniline
Acenaphthene
2,4-Dinitrophenol
4-Nitrophenol
Dibenzofuran
2,4-Dinitrotoluene
Diethylphthalate
4-Chlorophenyl-phenylether
Flourene
4-Nitroaniline
4,6-Dinitro-2-methylphenol
N-Nitrosodiphenylamine*
4-Bromophenyl-phenylether
Hexachlorobenzene
Pentachlorophenol
Phenanthrene
Anthracene
Di-n-butylphthalate
Flouranthene
Pyrene
Butylbenzylphthalate
3,3' -Dichlorobenzidine
Benzo(a)anthracene
Chrysene
bis(2-Ethylhexyl)phthalate
Di-n-octylphthalate
Benzo(b)flouranthene
Benzo(k)flouranthene
Benzo(a)pyrene
Indeno(l ,2,3-cd)pyrene
Dibenz(a,h)anthracene
Benzo(g,h,i)perylene
Cannot be separated from diphenylamine.
29
-------
TABLE 12. TCLP LIST OF VOLATILE ORGANICS
Volatile component
Regulatory Limits (u)
(mg/L)
Acrylonitrile
Benzene
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chloroform
1,2-Dichloroethane
1,1-Dichloroethylene
Isobutanol
Methylene chloride
Methyl ethyl ketone
1,1,1,2-Tetrachloroethane
1,1,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1,1,1-Trichloroethane
1,1,2-Trichloroethane
Trichloroethylene
Vinyl chloride
5.0
0.07
14.4
0.07
1.4
0.07
0.40
0.1
36
8.6
7.2
10.0
1.3
0.1
14.4
30
1.2
0.07
0.05
(a) Proposed toxicity characteristic contaminants and regulatory levels, FR Vol. 51, No. 114, June 13, 1986,
p. 21652
(b) Performance indicator for this program.
30
-------
TABLE 13. NUMBER AND TYPES OF QC SAMPLES
Location and
sampling
parameter
Waste feed
Furnace bed ash
and filter
fly ash
(2 streams)
Limestone feed
Flue gas
Analytical parameter
Metals
Vol w/lib scan
Semivolatiles w/scan
PCDD/PCDF
PCBs
Organic halogens
Inorganic halogens
Metals
Vol w/lib scan
Semivolatiles w/scan
PCBs
PCDD/PCDF
TCLP volatiles
TCLP metals, Cr(VI)
Organic halogens
Inorganic halogens
Metals
Calcium
Particulates
Metals
Semivolatiles w/scan
PCBs
PCDD/PCDF
Vol w/lib scan
Chloride
Program
samples
3
14
3
3
3
3
3
6
12
6
6
6
12
6
6
6
3
3
3
3
3
3
3
(3x3)
6
MS*
2
2
—
—
—
—
—
2
2
4
—
—
2
2
—
—
2
2
—
2
—
—
—
—
2
FB/ Method
Audit Blank
1
1
1
1
1
1
1
1
1
__
__
__
__
__
_
__
__
—
1
1
1
1
1
7 2
1 1
LCS
—
—
1
1
1
1
1
—
—
—
1
1
—
—
1
1
—
—
—
—
1
1
1
5
— "
Total
analyzed
6
17
5
5
5
5
5
9
15
10
7
7
14
8
7
7
5
5
4
6
5
5
5
23
10
* Matrix spikes for volatile organics contained carbon tetrachloride. TCLP matrix spiking was performed
on the generated leachate.
MS Matrix spike
FB Field Blank
LCS Laboratory control sample
31
-------
TABLE 14. ANALYTICAL QC MEASURES
QA measure
Minimum frequency
Method blank
Laboratory control sample
Calibration check sample
Matrix spike/matrix spike duplicate
Surrogate spike (GC/MS analysis)
Each sample set extracted, or daily for volatile
organics.
Each sample set extracted for VOST semivolatile
organics, PCBs, PCDD/DFs, organic and inorganic
halogens.
Daily for volatile organics, every 12 hours for
PCB/PCDD/PCDF/SV.
Each sample matrix for metals and volatile organics
(except VOST). Standard additions are used for M5
train digestates, chloride impingers, and TCLP
leachates.
Each sample.
Long-term corrections most often result in formalized changes to QA/QC procedures. Corrective actions
taken during this program are shown in Section 5 of this report, and in Appendix II of the field notes.
Audits
At the request of EPA, five audits were conducted for this program. Each audit was performed under
the guidance and supervision of EPA/RREL. Results of the audits are contained in Section 5 of this report.
The first audit consisted of a technical systems review of ATC's field sampling effort and OES's CEM
system. Representatives of S-CUBED, the audit contractor selected by EPA, were at the CBC facility throughout
the field sampling and process monitoring activities.
Three audits were conducted to observe and report on the laboratory operations. One audit was
performed by S-CUBED at the Clean Harbors Analytical Services Laboratory and focused on the analytical
procedures for volatile organics by Methods 5040 and 8240. Two audits were conducted by Research Triangle
Institute at Triangle Laboratories, Inc., which observed the analysis for semivolatile organic compounds by
Method 8270, and PCDD/PCDF analyses by Method 8290.
The final audit was an on-site VOST audit by S-CUBED, evaluated by Research Triangle Institute.
32
-------
SECTION 5
PERFORMANCE DATA EVALUATION
INTRODUCTION
As noted in the introduction to this report, the criteria used in this program to evsduate the efficiency of
the CBC in destroying hazardous constituents were the following:
o Destruction and removal efficiencies (DREs)
o Flue gas emissions (including acid gas removal, particulates, organics, metals, and CEM evaluations)
o Organics destruction
o Toxic metals distribution
o TCLP results
o Physical parameters
o System operating conditions
The primary objective of the demonstration was to gather data needed to evaluate the CBC technology
and to determine the feasibility of conducting a SITE Program Demonstration Test at the McColl Superfund Site
in Fullerton, California. The treatability study was limited by the permit conditions of OES's research facility.
The most significant permit restrictions were as follows:
o Federal and State RCRA RD&D permits — limited the concentration of sulfur in the waste feed to 5%.
This did not allow for the evaluation of the CBC system using McColl waste with a higher sulfur content.
o OES and San Diego APCD - restricted the McColl waste feedrate to 200 Ib/hr, (which equates to a total
throughput of waste and sand of 400 Ib/hr). This prevented observation of the system operation at or
near its full processing capacity.
o APCD permit — restricted the SO2 emissions to a not-to-exceed value rather that a time-weighted average.
This resulted in the use of large quantities of limestone to minimize the SO2 emissions.
The permit restrictions prevented optimization of the Ca/S ratios and evaluation of the CBC control
system's ability to react to intentionally induced abnormal operating conditions. The relatively high heating value
and variable sulfur content of the McColl feedstock could result in transient surges which the treatability study
had wanted to simulate. The SITE program was not involved in these permit decisions but nevertheless was
required to modify the test program in keeping with their limitations. In the event that a SITE Demonstration
Test is approved at the McColl Site, the test will be designed to elicit more detailed data than was required in
the treatability study.
The tables presented in the balance of this section are summaries of the more-detailed ATC tables included
in Appendix H of the field notes. Details of any table can be found in the Appendix II tables in the field notes,
as footnoted. Conversions of sample concentrations to emission rates, and concentrations of contaminant per
weight of feed are presented in the Appendix II Tables 3-14 through 3-20 in the field notes. The raw analytical
and QC data are provided in ATC's Volume II [1], not appended to this report.
33
-------
The numerals and symbols used in the tables are as follows:
o Numerical values - Concentrations are reported when the concentration of the analyte was greater than
the quantitation limit (the minimum concentration of the analyte that can be quantified using the specific
procedure). Quantitation limits (i.e., minimum detectable limits) are tabulated in Appendix II of the field
notes (ATC's Appendix A).
o Not detected (ND) - This designation indicates that the component was not identified in the sample.
o Less than (< ) - This symbol is used to show that the analyte was positively identified in the sample, but
that the quantity detected was less than the accepted limit for accurate quantitation.
PERFORMANCE CRITERIA RESULTS
Destniction and Removal Efficiency
The Test 3 unblended waste feed was spiked with sufficient quantities of CC14 to determine the destruction
and removal efficiency of the system while the waste was being processed. The DRE results for CC14 are shown
in Table 15. These results indicate a quite satisfactory DRE, indicating that the CBC performance in destroying
CC14 met the existing regulatory requirements of 99.99%.
Table 16 presents combustion emission concentrations and CBC efficiencies that were calculated for
benzene, toluene, ethylbenzene, and xylene (BTEX). The calculated emission rates for BTEX compounds ranged
from less than 2.1 mg/hr to 76 mg/hr. (See Appendix II, Table 3-14 of the field notes). These rates are
consistent with the expected formation of small quantities of PICs in combustion processes.
Flue Gas Emissions
Flue gas emission results are consolidated and summarized in Table 17. The table shows ranges (derived
from the laboratory report in Appendix H of the field notes) for volatile and semivolatile organics, PCBs,
PCDDs/PCDFs, metals, chlorides, and particulates.
TABLE 15. DREs FOR CC14
Parameters
CCl4(a)
Waste feed concentration (ppm)
Waste feed rate (Ib/hr)
Compound feed rate (Ib/hr)
Stack Flow (dscfm)
VOST volume (dsl)
VOST concentration (ng)
Emission rate (Ib/hr)
DRE (%)(b)
1168
200
0.224
946
19.633
77.5
1.42 E-05
99.9936
(a) Average values. See Appendix II Table 3-9 in Field Notes for details.
Compound feed rate - Emission rate
(b)
DRE =
Compound feed rate
-x 100
34
-------
TABLE 16. COMBUSTION EMISSION CONCENTRATIONS FOR SELECTED COMPOUNDS
Parameter
Waste feed concentration (ppm)
Waste feed rate Ub/hr)
Compound feed rate (Ib/hr)
Stack Flow (dscfm)
VOST Volume (DSL)
Vost Concentration (ng)
Emission Rate (Ib/hr)
CBC efficiencies (%)(b)
Benzene(a)
4.1
232
9.63 E-04
924
18.4
54
1.02 E-05
98.94
Toluene(a)
28
232
6.50 E-03
924
18.4
114
2.20 E-05
99.66
Eithyl
Benzene(a,c)
16.8
232
3.90 E-03
924
18.4
37
7.CI3 E-06
99.82
Xylene(a,d)
113
232
2.62 E-02
924
18.4
127
2.52 E-05
99.90
(a)
(b)
(c)
(d)
Average values for comparison purposes only. NOT TO BE USED FOR CALCULATION. See Appendix II Table 3-10 in
Field Notes for details.
CBC efficiency =
Compound feed rate - Emission rate
Compound feed rate
x 100
No ethyl benzene was detected in 8 of 9 samples. The quantisation limit was used as the basis for these
calculations.
No xylene was detected in 3 of 9 sample. The quantisation limit was used as the basis for these calculations.
Acid Gas Removal -
Chloride samples were collected for each test; the results are shown in Table 18. No detectable amounts
of chloride were found in the flue gas even though detectable amounts were analyzed in the waste feed.
Emissions rates were based on minimum detection limit concentrations and were calculated to average less than
0.0083 Ib/hr, well below the RCRA performance standard of 4 Ib/hr. This indicates very good chloride control
in the CBC.
Flue Gas Particulates —
The RCRA paniculate emissions standard for incinerator operations is 0.08 graims/dscf (corrected to 7%
Oj). The test results for the McColl waste showed paniculate loadings to be 0.0029 gr/dscf corrected to 7% O2
for Run 1, 0.0035 (Run 2), and 0.0023 (Run 3). As a comparison, emissions from state of the art equipment
typically range from 0.01 to 0.02 gr/dscf. The test results indicate that the baghouse collectors very adequately
captured stack gas particulates. However, the SCAQMD regulations require particulate emissions to be less than
0.002 gr/dscf corrected to 12% CO2 (approximately 0.0022 gr/dscf at 7%
35
-------
TABLE 17. FLUE GAS RESULTS
Parameter
Quantities (a)
Volatiles (ng)
SemivoIatiles (ug)
Hetals (ug)
Chlorides
Particulates
Carbon Tetrachloride . . . .
Benzene
Toluene
Ethylbenzene
Total xylenes
Phenol
Benzyl alcohol
Isophorone
Benzoic acid
Naphthalene
2-Hethylnaphthalene . . . .
Phenanthrene . .
Dimethylphthalate
Di-n-butylphthalate
Bis (2-ethylhexyl) phthalate
PCBs (ug)
PCDDs/PCDFs (ng)
Barium*
Cadmium*
Chromium*
Cobalt
Copper
Lead*
Manganese
Nickel
Tin
Zinc
Chloride (mg/l)
Emission Rate (lb/hr) . . .
Grain Loading (gr/dscf) . . .
Emission Rate (lb/hr) . . .
ND - 140
29 - 120
ND - 660
NO - 130
NO - 840
30 - 79
(b)
(b)
(b)
<10 - 16
(b)
ND
(b)
(b)
(b)
(c)
(b)
15 - 26
5.3 - 11
18 - 71
3.2 - 13
37 - 50
ND - 13
8.5 - 33
20 - 84
68 - 647
406 - 2970
ND
<0.0083
0.0029
0.0242
(a) Ranges are for comparison purposes only. NOT TO BE USED FOR CALCULATION. See Appendix II Tables 3-6, 3-11,
and 3-13 in Field Notes for details.
(b) Trace amounts. Laboratory contaminants.
(c) Trace Amounts. Incorrect Identification.
ND Not detected. See Appendix II in Field Notes for quantitation limit.
< Detected at levels less than the quantitation limit.
* Priority pollutant metals.
36
-------
TABLE 18. CHLORIDE IN THE FLUE GAS
Run number*
2
Sample concentration, mg/L
Sample volume, ml
Chloride collected, mg
Emission rate, Ib/hr
< 1.5
86.7
< 0.13
< 0.0090
< 1.5
72.5
<0.10
< 0.0077
82.5
< 0.12
< 0.0082
* Average values for comparison purposed only. See Appendix II Table 3-13 in Field Notes for details.
< Detected at levels less than the quantitation limit.
Flue Gas Organic and Metals Concentrations —
Volatile organics - Organics were sampled and analyzed not only to determine the degree of destruction
of organics initially present in the waste feed, but also to identify and quantify any new hazardous components,
not present in the waste feed, that were formed in the combustion process.
The volatile organic concentrations of benzene, toluene, xylenes, and ethylbenzene found in the flue gas
(Table 17) should be considered in light of PIC formations and VOST tube contamination. Benzene, toluene,
and xylenes, although present in low concentrations in the waste feed (Table 16), are also common PICs of
combustor processes, (i.e., can be created as a result of the combustion conditions). These aromatic compounds
also can be contaminants in the VOST train components. It was not determined whether the reported flue gas
organics concentrations (Table 17) are the result of incomplete thermal destruction, PIC formation, or sampling
train contamination. A combustion blank (feed sample prior to burning) was collected, and the organic
concentrations adjusted for background combustion products. Nevertheless, the flue gas volatile organics
concentrations are low.
Sernivolatile organics — Except for phenols, eight of the semivolatile compounds detected in the flue gas
(Table 17) have been discounted as sample medium contamination prior to or after sampling. The phenol
concentrations were low.
Inspection of the PCB analytical data shows trace concentrations of PCBs that were slightly above the
established limits of detection. Since PCBs and chlorinated hydrocarbons are not known to be contaminants at
the McColl Site and were not identified in the pretest feedstock sampling performed by ATC (Appendix II of the
field notes), it is questioned as to why PCBs were identified. The methodology used to analyze PCBs can
produce results that identify the spectra as PCBs when they could be something else. It is the conclusion of this
report that these reported materials are not PCBs.
The presence of trace amounts of PCDDs and PCDFs in the flue gas - while neither they nor chlorinated
hydrocarbons are suspected as site contaminants nor were detected in the waste feed, bai ash, fly ash, and pretest
feedstock samples -- indicates that the presence of these congeners can be attributed to sources other than the
McColl waste. The ATC report concludes that the compounds probably originated as laboratory contaminants.
37
-------
It is the conclusion of this report that the indicated PCDD/DFs are laboratory contaminants.
PICs — In general, PICs and other undesired combustion by-products appear to have been controlled in
the process.
Metals — Only low quantities of metals were found in the flue gas. The analytical methodology used to
detect metals in the flue gas cannot differentiate between those metals that are vaporized and later condense, and
those that are carried through the gas stream as discrete particles. The reported quantities are the combined
amounts of metal particles from both mechanisms.
Continuous Emissions Monitoring (CEM) Results -
The CEM results in Table 19 show that CO, NOX, and THC emissions were controlled within the OES
permit limitations. A momentary spike of THC resulted in an interlock shutdown of the system, interrupting
sampling for about an hour. During the test, the system operated at higher than usual excess air levels. The
effect of lower excess air on CO and THC emissions will be re-examined during the Demonstration Test.
An independent audit (Appendix n of the field notes) of the CO, NOX, and SO2 analyzers revealed that
the CO and NOX analyzers operated within performance specifications, but that the SO2 analyzer did not. The
SO2 emission results therefore were not quantifiable. It is likely that the SO2 emissions in the latter part of the
program were controlled to low levels as a result of the large amounts of limestone used to control the spikes (6:1
to 14:1 Ca/S ratio). It is expected that further testing under the SITE program will establish the optimum Ca/S
ratios needed to control SO2 emissions within environmentally acceptable limits. The audits also revealed that
the analyzers — although calibrated to manufacturer's specifications — were not calibrated in accordance with EPA
standard reference methods for multipoint calibrations. Consequently, calibration procedures were not in strict
conformance with prescribed EPA QA/QC protocol.
The OES CEM output data are displayed in three ways: 1) direct reading from the analyzer meters; 2)
tables, graphs, and a video display of major operating parameters printed by DAS software; or 3) strip chart
recordings. The DAS provides information at two-minute intervals for internal system records and video
displays, and at 15-min intervals for printed graphs and charts. Observations during the test of the three output
sources revealed that the analyzer readings, DAS outputs, and strip chart readings differed slightly. The average
DAS values were used throughout this report. The CO, NOX, and SO2 tabulations in the OES test results
(Appendix HI of the field notes) contain many "zero" readings. At low input concentrations analyzers often
record these concentrations and those less than the detection limits of the instrument as zero. Some degree of
error is inherent in the data; the amount of error for each parameter can be estimated to be about 7% for CO and
13% for NOX. (See audit report in Appendix II of the field notes). The SO2 data cannot be evaluated because
the SO2 analyzer did not operate within performance specifications.
Organics Destruction
Organics results are shown in Table 20. The test runs were performed with three different feed materials:
Run 1 with McColl waste blended with an equal weight of sand; Run 2, unblended McColl waste; and Run 3,
unblended McColl waste spiked with 6,000 ppm of CC14 as a performance indicator for DREs. Note that the
tables present ranges of concentrations for comparison purposes only. The baseline data for these tables can be
found in the footnoted references.
Volatile Organics —
Although the Run 3 feed was spiked to 6,000 ppm CC14, none of the feed sample concentrations exceeded
1,500 ppm. This can be attributed to several possibilities such as: high volatilization of the CC14 during transfer
into the feed bunker, or in the feed bunker itself (CC14 concentrations in the carbon filters were not evaluated);
nonhomogeneous mixing of the CC14 in the feed material (not likely since visual feed observation and precise
38
-------
TABLE 19. CONTINUOUS EMISSIONS MONITORING RESULTS
Test Period
3/28/89
10:55-19:12
Pretest Fine
Tuning
3/29/89
9:45-11:14
Test No. 1
11:14-16:33
Run No. 1
16:33-18:18
End Test No. 1
3/30/89
7:45-11:28
Test No. 2
11:28-14:52
Run No. 2
14:52-17:25
Test No. 3
17:25-21:23
Run No. 3
Permit Limits
In situ
Probe
C0(a) N0x(a) THC(a) S02 C02(a) 02(a) 02(a)(b)
(PPnO (ppm) Cppm) (ppm) (%) (%) (%)
0-45 0-20 0-7.4 (c) 4.7-9.9 9.5-14.1 9.1-12.8
22.4-39.9 22.5-55 4.9-12.4 5.3 ns
(a) Ranges for comparison only. NOT TO BE USED FOR CALCULATIONS. See Appendix III tables in Field Notes for
details.
(b) Wet probe. Readings are not correct for moisture.
(c) Not quantifiable. Sulfur dioxide analyzer was not operating within performance specifications.
(d) Zero values are calibration results.
39
-------
TABLE 20. ORGANICS AND HALIDE RESULTS
Parameter
Waste feed(a)
(mg/kg)
Fly ash(a)
(mg/kg)
Bed ash(a)
(mg/kg)
Volatiles
CC14
Benzene
Toluene
Ethylbenzene
Total Xylenes
ND - 1500
ND-7
14-49
ND-30
39 - 210
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Semivolatiles
Phenols
Benzyl alcohol
Isophorone
Benzoic Acid
Naphthalene
2-Methylnaphthalene
Phenanthrene
Dimethylphthalate
Di-n-butylphthalate
Bis(2-ethylhexyl)phthalate
PCBs
PCDDs/PCDFs
ND
ND
ND
ND
25-33
22-46
<10
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
(b)
(b)
(b)
ND
ND
ND
ND
ND
ND
ND
ND
ND
(b)
(b)
(b)
ND
(b)
Halides
(a)
(b)
ND
Organic chloride
Inorganic chloride
ND - 540
ND - 240
Ranges are only for comparisons of concentrations in
MATERIAL BALANCES. See Appendix II Tables 3-1,
Trace amounts. Laboratory
contamination.
Not Detected. See Appendix n in Field Notes for quantii
ND
83 - 530
solid streams. NOT TO
3-2, and 3-4 in Field Notes
tation limits.
ND
ND
BE USED FOR
for details.
Detected at levels less than the quantitation limits.
40
-------
analytical results indicate good mixing); nonrepresentative samples; or errors in preparing the spiking quantities.
It is not clear where the loss of CC14 occurred.
Table 20 indicates that no organics present in the waste feed were found in quantifiable concentrations
in the ash. This indicates that the CBC successfully destroyed the organic contaminants.
Semivolatile Organics —
Analyses of the bed ash and fly ash samples indicated that trace amounts of three semivolatile plasticizers
were present. Insofar as these components were found in equivalent concentrations in the laboratory blanks, their
presence is attributed to laboratory contamination.
No PCBs were found in the waste feed or ash streams. One trace PCDD/DF congener in one bed ash
sample has been attributed to laboratory contamination. PCB, PCDD, and PCDF compounds have not been
detected as contaminants at the McColl Site.
Tentatively identified compounds (TICs) are those compounds whose spectra closely — but not exactly
— fit the spectra of known compounds, and whose identity is assumed but cannot be verified. These compounds
are listed in Appendix II Table 3-8 of the field notes.
At the request of DHS, EPA added to the program the analysis of polynucleair aromatic hydrocarbons
(PAHs), specifically in the fly ash and bed ash streams, to determine further if the ash had been detoxified. The
results of the matrix spike analyses are presented in Appendix II Table 4-14 of the; field notes and will be
evaluated further by DHS.
Halides —
Organic chloride was detected only in the Run 3 sample, and is attributed to the CC14 material used to
spike the Run 3 waste feed.
Toxic Metals Distribution
Total metals results are shown in Table 21 The limestone stream is included in the table since limestone
interacts in the process in several ways: as a source of metals; as a matrix in which metals can be complexed
or organics absorbed; and as a pH adjuster which affects the chemistry of some of the contaminants.
Limestone was used in large amounts to control the SOtj emissions and, as a result, the ash material had
a high concentration of lime. It can be seen from the ranges in the table that metals were distributed mostly in
the fly ash and bed ash. Metal concentrations in the fly ash were consistently higher than in the bed ash —
probably due to fly ash particles being finer, which provide more surface area for metals adherence. The metals
concentrations found in the solid streams were low.
TCLP Results
TCLP Metals -
Although the total metals results give little indication as to the mobility of the metals in the ash, the TCLP
leachate results (Table 22) allow a much narrower interpretation of the data. The leachate amounts of the detected
metals were all well below the regulatory limits.
TCLP results are used as a criterion to determine if a waste is hazardous. However, once a waste is
defined as hazardous, it continues to be hazardous until it is delisted. The delisting criteria are established on
a case by case basis. Various dispersion/leaching models are used to determine whether or not ash resulting from
a thermal process will release toxics into the environment. In the past, the delisting criiteria for other processes
41
-------
TABLE 21. TOTAL METALS RESULTS
Parameter
Antimony
Arsenic*
Barium*
Beryllium
Cadmium*
Calcium
Chromium*
Cobalt
Copper
Lead*
Manganese
Mercury*
Nickel
Selenium*
Silver*
Thallium
Tin
Zinc
Waste feed(a)
(mg/kg)
ND
ND
65 - 136
ND-0.7
ND
NA
45-65
3.6-6.8
14-95
ND
95-211
ND
15-19
ND
ND- 1.1
ND
ND-6.3
33-49
Limestone(a)
(mg/kg)
ND
ND
13-14
ND
0.25
351,000-382,000(b)
0.9-1
ND
0.6- 1.1
ND
44-58
ND
ND - 1.0
ND
ND - 1.0
ND
ND
1.2-4.4
Fly ash(a)
(mg/kg)
ND
ND
44-49
ND
ND - 0.20
NA
17-26
1.4- 1.7
5.1-9.1
ND
114-148
ND
5.1-9.2
ND
ND
ND
ND
6.8 - 23
Bed ash(a)
(mg/kg)
ND - 5.0
ND
11 - 16
ND
ND - 0.45
NA
4.4 - 7.4
ND - 0.75
1.5-2.5
ND
35-41
ND
ND-4.1
ND - 3.3
ND - 0.99
ND
ND
6.1-9.4
(a)
(b)
NA
ND
*
Ranges are only for comparison of metals concentrations in solid streams. NOT TO BE USED
FOR MATERIAL BALANCES. See Appendix H Tables 3-1, 3-3, and 3-4 in Field Notes for
details.
Calcium content of the limestone was measured to range from 35 to 38%.
Not analyzed
Not detected. See Appendix II in Field Notes for quantitation limits.
Priority pollutant metals
42
-------
TABLE 22. TCLP RESULTS
Parameter
Fly ash(a)
(mg/1)
Bed ash(a)
(mg/1)
Regulatory
limit(b)
(mg/1)
Volatiles
Toluene (c)
ND- < 0.025
ND
14.4
Metals
(a)
(b)
(c)
ND
NL
Arsenic
Barium
Cadmium
Chromium
Chromium VI
Lead
Mercury
Selenium
Silver
See Appendix II Table 3-5 in
Regulatory limits proposed in
No other volatile compounds
ND
0.10-0.30
ND
0.07 - 0.08
0.04 - 0.05
ND
ND
ND - 0.07
ND
ND
0.10-0.20
ND
0.04 - 0.05
0.02 - 0.04
ND
ND
ND
ND
5.0
100
1.0
5.0
NL
5.0
0.2
1.0
5.0
Field Notes for details.
Federal Register
were detected.
Not detected. See Appendix II in Field Notes
Not listed in proposed regulat
ory reference.
Vol. 51, No. 114, June 13, 1986.
for quantitation limits.
Detected at levels less than the quantitation limit.
43
-------
have established levels more stringent than the criteria that defined the waste as hazardous; The leachate levels
from this test were very much lower than the TCLP regulatory limits. It is likely, therefore, that the ash from
the CBC could be delisted successfully. In addition, if the McColl fly ash and bed ash can pass the TCLP test
based on the leachate performance of the ash versus the regulatory standards, the ash can be defined to be
nonhazardous. Since the treatability study results prove that the ash meets this criterion, the on-site
Demonstration is expected to produce ash that is nonhazardous. However, since the TCLP test is sensitive to
the pH of the matrix, the on-site program must determine whether or not the TCLP results can be replicated when
the Ca/S ratios are optimized, the limestone use is reduced, and the pH of the material changes.
TCLP Organics -
Toluene (at a concentration of < 0.025 mg/L) was the only organic compound detected in the TCLP
leachate, indicating that essentially no leachable organic contaminants remained in the fly ash and bed ash.
Physical Parameters
Physical parameters are listed in Table 23. Note the expected increase in ash percentages, pH, and bulk
density as compared to the waste feed stream. No out-of-the-ordinary physical results were found in the residue
streams.
System Operating Conditions
The following is a summary of the operational events that caused process interruptions.
Feed of waste material began on March 28, 1989 at approximately 11:00 a.m. with the SO2 analyzer
interlocks set at 200 ppm to alarm, 250 ppm to interlock. After several hours, several SC^ spikes occurred. As
TABLE 23. SUMMARY OF PHYSICAL PARAMETERS
Parameter
Waste feed(a)
Fly ash(a)
Bed ash(a)
Moisture (%)
Ash (%)
pH
Bulk density (Ib/cu ft)
Heating value (Btu/lb)
Sulfur (%)
Carbon (%)
Hydrogen (%)
Nitrogen (%)
TOC (%)
8.34
72.6
2.3
57.9
1253
4.4
7.6
1.6
0.1
7.2
< 0.001
98.2
12.6
76.8
—
3.6
—
—
—
—
< 0.001
99.2
12.1
88.3
1.0
—
—
—
—
(a) Average of three runs. See Appendix II Table 3-7 in Field Notes for details.
44
-------
the waste feed rate increased, the limestone feed also was increased to enhance SQ, capture. The installed
limestone feed auger was found to be undersized.
Feed was interrupted for about an hour to install a larger auger. (Several limestone augers had been
calibrated prior to the test.) By 19:30, after several hours of system progression to full operating conditions, the
waste feed auger proved unable to deliver feed at the target feed rate. The feed auger was removed and replaced
during the night with a larger auger. System operation continued early on the morning of March 29. Waste feed
recommenced at approximately 08:00. Waste feed and limestone feed were quickly increased to the established
test conditions. Several SO2 spikes occurred with momentary feed interruptions, proving that the SO2 interlocks
were operating. Sampling for Run No. 1 began, but was interrupted due to a THC spike that interrupted
sampling for about one hour (also proving that the THC interlocks were operating), Feed and sampling
recommenced; additional SO2 spikes caused additional momentary shutdowns.
By early evening it was decided to lower the SO2 interlock settings to 75 ppm alarm/100 ppm interlock.
OES chose the addition of excess limestone to prevent feed interruptions caused by SO2 spikes. This precluded
the study of optimum Ca/S ratios. Runs Nos. 2 and 3 were performed on March 30 with only a few interruptions
caused by SO2 spikes. The CC14 spiked material initially was difficult to feed, but this was resolved by a change
in the feed procedure.
In general, the system components operated without major interruptions, allowing the sampling and
analysis effort to proceed without problems. Except for the SO2 analyzer, process instrumentation performed
reliably.
The facility.permit limitations prevented the observation of system response to variations in the feedstock
and the evaluation of SO2 emissions, as the Ca/S ratios were varied. These parameters would be evaluated fully
in any on-site Demonstration.
QUALITY ASSURANCE/QUALITY CONTROL
The quality assurance/quality control protocols for this program were detailed in die Test Plan. Changes
to the Test Plan QAPP are discussed in this section and in Appendix II of the field notes,
Data Quality Objectives
Precision, accuracy, and completeness objectives were established in the QAPP to evaluate the quality
of the collected data.
Precision is the measurement of agreement of replicate results without consideration of the true results,
and is expressed as the relative percent difference (RPD). Therefore, the closer the RPD value is to zero, the
more reliable are the data. Accuracy is defined as the nearness of a result to an accepted reference or true value,
and is expressed as percent recovery. The closer a percent recovery value is to 100%, the more reliable the data.
Completeness is an estimate of the percentage of measurements made whose values are judged to be valid as
compared to the amount of expected data. Completeness values close to 100% indicate reliable data. Note that
completeness percents are subjective evaluations of ATC, based on experience and laboratory knowledge.
Percent recovery ranges were established for each of the analyses, thus quantitatively determining the
reliability of the data. Table 24 lists the percent recovery objectives for each analytical parameter, and the results
of the QC analysis. A discussion of the reliability of the data based on accuracy, precision, and completeness
objectives follows.
45
-------
TABLE 24. PERCENT RECOVERIES RESULTS
Parameter
MS for metals-waste feed
limestone and ash
MS for metals -flue gas
MS and LCS samples for halides
MS for metals-TCLP leachates
SS for volatile organics-waste
feed and ash
SS for volatile organics-TCLP leachate
MS for volatile organics-TCLP leachate
SS for volatile organics-VOST tubes
LCS for volatile organics
SS for SV organics-waste feed and ash
SS for SV organics-MM5 train
LCS for SV organics
MS for SV organics-ash
SS for PCBs
LCS for PCBs
SS for PCDD/DF-waste feed and ash
SS for PCDD/DF-MM5 train
LCS for PCDD/DF
Appendix II*
table
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9
4-10
4-11
4-12
4-13
4-14
4-15
4-16
4-17
4-18
4-19
Established
% recovery
objectives
10-120%
50-150%
75-125%
50-120%
50-150%
50-150%
50-150%
50-150%
50-150%
10-120%
10-120%
10-135%
—
40-120%
50-140%
40-120%
40-120%
50-120%
% Completeness of
objectives
100%
100%(a)
100% (b)
100%
97%
100%
100%
100%
100%
80%
80%
95%
(c)
98%
97%
85%
85%
98%
(a) Tin and zinc were not evaluated.
(b) Iodide was not evaluated.
(c) QA objectives for this analysis were not established for this program.
LCS Laboratory control spike
MS Matrix spike
SS Surrogate spike
* Appendix n Tables in Field Notes
46
-------
Flue Gas Samples
Acid Gas Removal —
Halide recoveries were all within the 75-125% range established for the program. This indicated good
reliability of the data, except for iodide, which was not included in the analysis.
Organics and Metals —
VOST sample recoveries for volatile organic surrogate spikes ranged from 80 to 110% and, for the
laboratory control samples, from 80 to 118% - all well within the 50-150% established range.
Recoveries for PCBs, PCDDs, and PCDFs were near mid-range for the MM5 samples, indicating valid
data.
Percent recovery objectives for the flue gas metals were established at 50-150%, but the actual values
were 80-115%. This indicated good recovery and reliability of the data. Tin and zinc results were not calculated
because the spiking level was below the native concentration of the analyte. However, since the recoveries of
these two metals showed relatively high concentrations, the data appeared to be reliable - based on the accuracy
and precision of the analytical method used to determine their concentrations.
Organics Samples
Volatile Organics —
The waste feed, bed ash, and fly ash recoveries for volatile organic surrogates ranged mostly between
80-130%, except for three results, which exceeded the 150% established objective. Since they exceeded the
objectives, and all other values were well within the objectives, the volatile organics data are judged very reliable.
Semivolatile Organics —
Accuracy of the SV data was evaluated through analysis of surrogate compounds spiked into all samples.
Also, ash matrix spike pairs and LCSs for the waste feed and MM5 train samples were analyzed along with the
program samples.
A high outliers percentage recovery in the bed ash required re-extraction and re-analysis of the sample.
Outlier percentages still remained high. The troublesome recoveries were in the phenol and 2-fluorophenol
surrogates, indicating that phenolic data should be evaluated cautiously. It should be noted that ATC found the
recovery of semivolatile surrogates typically troublesome, not only in the samples in this program, but in many
other programs as well. The other semivolatile recoveries indicate that the balance of the semivolatile data are
quite good.
Two pairs of ash matrix spikes were prepared by the addition of a solution containing 16 polynuclear
aromatic hydrocarbons (PAHs). Although no QA objectives were established in the Test Plan, the resultant
recoveries were well within typically reported ranges for reagent water.
Recoveries for PCBs in the waste feed and ash samples indicated that recoveries for the waste feed were
about mid-range and, for the ash streams, very close to the lower objective (with one outlier in the bed ash).
This suggests that the reliability of PCB data is somewhat weaker for the ash streams than for the waste feed
stream, although all data can be judged valid and within the QC objectives.
Analysis of the samples for PCDDs and PCDFs resulted in recoveries well above mid-range for the waste
feed and ash samples. Although completeness was estimated at only 85% — less than the 90% objective of the
Test Plan — the data are considered very reliable, since all outliers were biased high.
47
-------
Toxic Metals Results
Percent recovery objectives for the waste feed, limestone feed, fly ash, and bed ash metals were
established at 10-120%, but the actual values lay between 60-110%. This indicates a good recovery and strong
reliability of the metals data.
TCLP Results
TCLP Metals -
The recovery objectives for the TCLP metals were established at 50-120% but actually ranged between
80-105%, indicating very good reliability of the TCLP metals data.
TCLP Organics -
Organic TCLP leachate recoveries for volatile organic surrogate spikes ranged from 86 to 107% and, for
TCLP leachate volatile organics, matrix spikes from 100 to 120% - also indicating very reliable data.
Deviations from the OAPP
Modifications to the QAPP resulting from field changes, program oversights, or sampling/analysis
requirements are discussed in ATC's test report in Appendix II of the field notes, and are listed below for
reference.
The initial sampling schedule shown in the Test Plan was changed from 3 days to 2 days because less
feedstock was available than originally anticipated, thereby shortening the test period. Also, the CEM data to
be taken after the spiked feed run were deleted from the program because the SO2 analyzer was not operating
within performance specifications.
Spiking concentrations of CC14 originally were established at 2,000 ppm, then increased to 3,000 ppm
to ensure sufficient concentrations to achieve calculated results of at least 99.99 %, but, due to a calculation error,
ended up at 6,000 ppm. This error inadvertently aided the DRE calculations since CC14 concentrations in the
waste feed samples were much lower than anticipated.
Two matrix spike duplicates of surrogate spiked PAHs were added to the chemical analysis program to
determine PAH recoveries in the ash stream samples. Analysis of the ash for PAHs was added to the program
to determine if the fly ash and bed ash material had been detoxified of residual PAHs.
One limestone feed sample was scheduled for each of the three runs, but was only taken for two runs -
the result of a field oversight. This oversight did not negatively affect the program, since the results from the
limestone samples for Runs 1 and 3 were nearly identical.
Field bias blanks were programmed for metals and MM5 semivolatile train QC, but they were
overlooked. Instead, reagent blanks were used. The laboratory audit report determined that this procedural
change had not significantly affected the data quality.
Four pairs of VOST tubes were analyzed for the third (spiked) run, instead of the 3 in the QAPP. This
was done to insure that adequate VOST samples were available for DRE calculations.
The QAPP ash sampling procedure was modified during the test because the originally planned method
- to penetrate the drums with a thief sampler - could not be executed as planned. Instead, for Run 1 the ash
was vacuumed into another container, and grab samples taken. For Runs 2 and 3, grab samples were taken
during the run. This modification provided a safer and more effective sampling procedure.
48
-------
The number of feed samples taken for Run 3 was increased from 2 to 8, to provide more representative
samples of the spiked waste. Both visual evaluation and waste feed sample analysis indicated that the spiked
material was adequately mixed in the feedstock.
Waste feed samples were prepared for semivolatiles analyses in accordance with the sonication procedures
of the Test Plan. However, due to the presence of elevated levels of hydrocarbons in the extract, an extra
dilution was necessary prior to analysis by GC/MS. Since this dilution would have made the surrogate
compounds undetectable, a re-extraction was performed using a 1-gram aliquot of the samples. This permitted
quantitation of the surrogate-spiked compounds.
Because of a laboratory error, the ash samples were prepared by soxhlet extraction instead of the
sonication specified in the Test Plan. No corrective action was taken because both methods are cited in SW846.
This procedural change did not negatively affect the generated data.
Audit Results
The audit reports have been placed in Appendix II of the field notes. Audit results are summarized
below.
CEM and Field Sampling -
EPA solicited the services of S-CUBED, San Diego, California, to be on site during the entire treatability
study test period. All field work performed by ATC was judged satisfactory, although the overall audit was given
a conditional rating because of problems relating to OES's SO2 analyzer.
The technical systems review audit initially revealed that the CEM systems were not calibrated using
multi-point gas calibrations. Instead, OES calibrated the analyzers using a zero gas, a single span gas, and an
electronic linearity check (essentially a single point gas calibration technique). This digression from the EPA
standard method for multipoint calibration was considered a minor concern, pending results of the analyzer
performance audit using NBS-traceable calibration gases.
The audit revealed that the CO analyzer was operating within performance specifications. The NOX and
SO2 analyzers were outside performance specifications. The THC analyzer was not audit<2d. On the evening of
March 29, 1989, OES changed out the NOX analyzer, which later was audited and found to be within performance
specifications. The SO2 analyzer was not changed out and was confirmed not to be operating within performance
specifications, a major concern. The audit report (attached to Appendix II of the field notes) concludes that"....
the SO2 measurements collected during the test burn are subject to an undefined degree of negative bias, which
will limit the application of the data:" This resulted in unquantifiable SO2 emissions data..
No field bias blanks were collected for the M5 and MM5 sampling trains. This was not discovered until
after the sampling effort had concluded. Instead, samples of all the field reagents were collected as a means to
measure reagent contamination. It was the audit's conclusion that the data quality was retained within acceptable
levels by use of this alternate technique.
Clean Harbors Analytical Services —
This review by both S-CUBED and EPA/RREL focused on the analysis of volatile organics. Method
performance was given a satisfactory rating.
Triangle Laboratories, Inc. —
TLI was audited twice — once for the analysis of semivolatile organic compounds, and once for analysis
of PCDDs/PCDFs. Both reviews achieved a rating of satisfactory.
49
-------
The first audit was conducted hi April 1989 by Research Triangle Institute (RTI), and focused on the
seraivolatile and PCB analyses. The audit was conducted over a 3-day period and reviewed project organization,
sample custody, sample extraction procedures, standards preparation, sample analysis, and QC. The one major
concern — that the MM5 blanks had not arrived with the rest of the samples — was corrected a week later when
the field blanks were received.
The audit concluded that the semivolatile analysis was satisfactory, and that the minor concerns would
not prevent achievement of the data quality objectives established in the QAPP.
In May 1989 a second audit was conducted by RTI that focused on the dioxin analysis. Only one minor
concern was expressed, giving the audit a rating of satisfactory.
Industrial Testing Laboratories -
No audits were conducted at ITL.
VOST-
Research Triangle Institute was solicited to conduct a field/analytical audit of ATC's VOST methods.
The audit conclusions are not yet available.
Summary
The QA/QC sample results indicated that reliable data had been collected during the test sampling and
that the deviations from the QAPP had not negatively affected the data quality. The audit results showed that,
except for the CEM analyzer calibrations and the deficiency in the SO2 analyzer, the methods and procedures
established hi the QAPP for the test program had been followed, and the data quality could be expected to be
satisfactory.
50
-------
SECTION 6
TREATABILITY STUDY COSTS
The demonstration project costs were calculated by Ogden Environmental Services (OES) and EPA. OES
presented the costs to prepare for and execute the treatability study at their pilot-scale facility. The EPA totalled
the costs for waste preparation, transportation, sampling and analysis, waste disposal, and report preparation.
These costs are not to be used to extrapolate to the potential on-site costs of a full-scale operation. Such on-site
costs will be determined during the on-site Demonstration Test.
The principal categories used to evaluate a typical on-site Demonstration, and their relationship to this
treatability study are:
o Site preparation costs — not applicable, since the treatability study was conducted at an existing pilot-scale
research facility.
o Permitting and regulatory costs — enabling OES to acquire test-specific permits for the facility.
o Equipment costs — not applicable, since the facility was an established, operating unit. Note that waste
pretreatment and posttreatment costs also were not included.
o Start-up and fixed costs - not applicable. This category is directed at on-site operations. The costs for
the facility modifications applicable to this study are included in the "facility modifications, repair, and
replacement costs" category shown below.
o Labor costs - providing the outside (contract) labor and in-house (OES) labor required to prepare the
system for operation, and to conduct the test.
o Supplies and consumables costs — included in this study.
o Effluent treatment and disposal costs — not applicable, as the CBC is a dry system and does not generate
liquids that require treatment and disposal.
o Residuals and waste shipping, handling, and transport costs — allowing the disposal of the residual ash.
o Analytical costs — EPA-supplied sampling and analysis. OES internal analysis costs are not included.
o Facility modification, repair, and replacement costs — OES-incurred in preparing the facility for the test.
o Site demobilization costs — providing only decontamination of the system.
The costs associated with each of the categories are shown in Table 25.
51
-------
TABLE 25. TREATABILITY STUDY COSTS
Cost
category
Site preparation
Permitting and regulatory
APCD fees
OES labor
Equipment
Start-up and fixed costs
Labor
Contract labor
OES labor
Supplies and consumables (a)
Effluent treatment and disposal
Residuals and waste shipping,
handling, and transport
Analytical costs
OES test
preparation
costs
N/A
$40,345
16,416
N/A
N/A
9,918
6,000
N/A
NI
NI
OES Test
execution
costs
N/A
N/A
N/A
29,298
36,416
13,430
N/A
NI
NI
EPA
costs
N/A
N/A
N/A
N/A
197,110
70,700
N/A
11,000
324,190
Facility modification, feed and
vent system modifications
Contract labor
OES labor
Site demobilization
(decontamination)
Supplies
Contract labor
OES labor
12,538
4,150
2,000
$91,367
2,594
8,400
12,000
$102,138
17,000
N/A
$620,000
TOTAL COSTS
$ 813,505
(a) Limestone costs were: $166/ton Pfizer limestone; $18.75/ton (1986 quoted price) Colton limestone.
N/A Not applicable
NI Not included
52
-------
SECTION 7
CONCLUSIONS
The objective of the demonstration was to evaluate the OES CBC technology for possible further testing
at the McColl Superfund Site under the SITE Program.
The treatability study results indicate that the CBC technology can be further evaluated under the SITE
Program. This observation is based on the results of the treatability study conducted in March 1989 at the OES
pilot-scale research facility in San Diego, California The treatability study data indicate that the CBC technology
can be used to destroy McColl Site organic contaminants while producing an ash residue that is nonhazardous.
Emissions of criteria pollutants can be controlled to regulatory levels. The test data are discussed below. Where
data is incomplete, extrapolations are based both on good engineering judgment and on comparisons to the
fully-operational OES 36-inch diameter CBC. (See Table 26.)
The following were not evaluated in this program: material and energy balances; equipment reliability;
labor, energy, supply, and maintenance requirements; economics; and pretreatment and posttreatment of waste
feed and residue streams. Of these, pretreatment of the waste feed, and ash pretreatment and posttreatment
requirements will most influence the final evaluation of an on-site operation. It is anticipated that some form of
waste pretreatment may be desirable to allow the CBC system to reliably process the McColl waste. Based on
TCLP results it is anticipated that no posttreatment at the McColl Site will be required. These considerations will
be evaluated in a more detailed SITE Demonstration Test.
TEST RESULTS
Destruction and Removal Efficiency
The destruction and removal efficiencies (DREs) for the CC14 performance indicator met the 99.99%
RCRA requirement. This indicates that the system is effective in destroying many organic contaminants.
Flue Gas Emissions
Acid Gas Removal —
Chloride was not detected in the flue gas. Using the quantitation limit (1.5 mg/1) for calculations, the
average chloride emission rate was calculated at less than 0.0083 Ib/hr, well below the RCRA performance
standard of 4 Ib/hr. Chloride emissions from an on-site demonstration are expected to fall well within existing
regulatory limits.
Flue Gas Particulates —
Particulate concentrations in the stack gas averaged 0.0029 gr/dscf (at 7% O^); the RCRA limit for trial
burns is 0.08 gr/dscf (at 7% O^). This indicates high efficiency capture of particulates in the CBC baghouse
filter, resulting in particulate emissions much lower than normal permitted levels. These low results have been
replicated in other tests [4,5]. However, SCAQMD regulations require that particulate emissions not exceed
0.002 gr/dscf, corrected to 12% CO2 (approximately 0.0022 gr/dscfm at 7% O^. Further testing of the CBC
is indicated to ensure compliance with SCAQMD emission limits.
Flue Gas Organics and Metals —
Low concentrations of some volatile organics were detected in the flue gas. Of those detected, almost all
are either common PICs or ubiquitous sample contaminants. Reported quantities of PCBs, PCDDs, and PCDFs
are attributed to laboratory contamination. PIC concentrations in the flue gas were insignificant.
53
-------
TABLE 26. COMPARISON OF CBC PARAMETERS TO OTHER TESTS
Swanson River,
Parameter Alaska(a)(d)
CBC diameter (in.) 36
Feed rate (lb/hr) 9000
CO (ppm) 8.7-17.5
O2 (%) 3.4-6.3
CO2 (%) 8.6-8.9
NOX (ppm) 82-95
SO2 (ppm) 19
THC (ppm) 2
Chlorides (lb/hr) 1.08-1.57
Particulates
(g/dscf @ 7% ©2) 0.0065-0.0190
DREs (%) > 99. 9999
(a) Soil contaminated with PCBs.
(b) Soil contaminated with No. 6 fiiel oil.
Stockton
California (b)(e)
36
4000
23.6-28
13.6
6.6-7.0
6.7-7.4
12-24.2
<2
Not measured(h)
0.045-0.046
99.9996
Pilot facility,
California(c)
16
200
0-51
8.3-19.6
2.3-13.7
0-67
(g)
0-84
< 0.0077-0.0090(1)
0.0029
99.9936
(c) Soil contaminated with residues from aviation gasoline manufacture.
(d) Reference [4].
(e) Reference [5].
(f) Results of treatability study.
(g) Not quantifiable. SO2 analyzer was not
(h) No chloride in feed.
(i) Based on minimum detection level. No
operating within performance
chlorides were detected.
specifications.
54
-------
Low concentrations of metals were also detected in the flue gas. Metals concentrations in the flue gas
are a combination of vaporized/recondensed concentrations, and discreet particles carried through the system.
Health risk assessments of an on-site demonstration would determine the importance of the flue gas metals
concentrations.
CEM Results -
The CO, NOX, and THC data indicated that the CBC can be operated and maintained within permit
limitations. It is expected that CO, NOX, and THC emissions of the on-site operation will change as the O2
performance is optimized, but other test data [4,5] indicate that emissions of a full-scale system also are low.
It is expected that the CO, NOX, and THC emissions can be controlled within existing SCAQMD limitations.
SO2 data cannot be evaluated since the SO2 analyzer was not operating within performance specifications.
The data indicate that sufficient quantities of limestone can be used to limit SO2 emissions, and it is expected the
SO2 emissions from a full-scale operation could be controlled to permitted levels. The SITE Demonstration
Program would determine the Ca/S ratios needed to achieve acceptable SO2 emissions.
Organics Destruction
Volatile organics present in the feed were destroyed in the solids streams. PCBs, PCDDs, and PCDFs
were not present in detectable concentrations in the waste feed, and were not found in the ash streams. The trace
amounts reported by the laboratory have been attributed to laboratory contamination. Since there are no
chlorinated hydrocarbons in the site waste, the Demonstration Test is not expected to produce any PCDDs or
PCDFs.
PIC concentrations and concentrations of other undesired combustion by-products were examined in the
treatability study. The CBC incineration of McColl waste can be expected to create only low concentrations of
PICs.
Toxic Metals Distribution
Metals were distributed mostly into the fly ash and bed ash streams. Low concentrations were detected
in the flue gas. Ash leachate tests showed the ash to be non-hazardous. The effect of the flue gas metals
concentrations would be evaluated in health risk assessments of an on-site Demonstration.
TCLP Results
TCLP Metals --
The TCLP leachate tests performed on the ash streams indicated that the McColl Site ash residue would
likely be classified as nonhazardous. However, since the TCLP test is sensitive to the pH of the tested matrix,
it is possible that the leachate results will change if Ca/S ratios are optimized and the amounts of limestone
reduced, thereby lowering the pH of. the matrix. Further testing of the system and optimization of limestone use
are indicated.
TCLP Organics -
Toluene was the only organic detected in the TCLP leachate; it was present at less than 0.025 mg/L in
only one of three samples. This indicates that the ash leachate is essentially organic-free.
55
-------
Physical Parameters
No out-of-the-ordinary physical characteristics were found in the ash and flue gas residues.
System Operating Conditions
Permit Limitation Effects —
The optimum Ca/S ratio required to limit SO2 emissions could not be evaluated because of permit
limitations. One important feature of the CBC technology is its capability to control SO2 emissions using dry
materials, rather than a wet scrubber. The necessity of using a wet scrubber to meet SO2 emission levels would
influence EPA's interest in conducting a Demonstration Test of the CBC at the McColl Site.
The permit limitations also affected the evaluation of the CBC control system's effectiveness in reacting
to intentionally-induced abnormal operating conditions. It is important that any on-site, full-scale system control
emissions with feedstock that varies widely in heating value and in SO2 content.
Waste feed was restricted to 200 Ib/hr feed rate and 5% sulfur content. This prevented evaluation of the
system's response to higher feed rates at the typically higher sulfur content of the McColl waste.
Process Interruptions —
The installed limestone feed auger was undersized. This was resolved by an auger change.
A larger waste feed auger was installed during the overnight standby operation.
SO2 system interlocks initially caused repeated feed interruptions. This was resolved by adding large
amounts of limestone, which precluded the study of Ca/S ratio optimization.
Operational Concerns —
The CO and NOX analyzers operated within performance specifications. The relative accuracy was within
7% and 13% respectively. The SO2 analyzer had electronic malfunctions that affected the reliability of the SO2
data. EPA determined that the SO2 data were not quantifiable.
The analyzer output data, strip chart recordings, and DAS outputs differed slightly. The discrepancies
were less than ±3% of full scale.
The CC14 spiked material initially was difficult to feed. This problem was overcome by removing the
feed mechanism and directly adding the spiked material.
OA/OC
The QC sample analyses indicated that reliable data were obtained from the study.
Costs
OES total costs for conducting the treatability study were $193,505 — based on permitting, labor,
supplies, facility modifications, and decontamination. EPA costs were $620,000 for waste preparation,
transportation, sampling and analysis, waste disposal, and report preparation. These costs cannot be extrapolated
to a full-scale system since they lack important cost elements of an on-site operation. It would be an objective
of the SITE Demonstration Test to determine full-scale operating costs.
56
-------
RECOMMENDATIONS
It is recommended that the CBC technology be evaluated further under the SITE Program. The SITE
Demonstration Test should be designed to test the CBC at its full operating capacity. The viability of this opinion
is dependent upon further deliberations among the regulatory agencies, discussions with the technology developer,
and continued evaluation of the treatability study results and other available test data to verify that die CBC can
meet the regulatory standards.
57
-------
-------
REFERENCES
1. Alliance Technologies Corporation. 1989. Sampling and Analytical Results foir the Preliminary Test
Burn Treatability Study of McColl Superfund Waste in Ogden's Circulating Bed Combustor Plant in San
Diego, CA, Volume II - Appendices, Alliance Project No. 7-536-999, Bedford, Mass.
2. GA Technologies, Inc. 1985. Process Demonstration Test Report for Trial Burn of PCB-Contaminated
Soils. GA-C18051. 10955 John Jay Hopkins Drive, San Diego, California
3. Journal of Air Pollution Control Association (JAPCA). 1987. Evaluation of a Pilot Scale Circulating
Bed Combustor as a Potential Hazardous Waste Incinerator. JAPCA, March 1987, Vol. 37, No. 3.
4. Ogden Environmental Services, Inc. 1988. Process Demonstration Test Report for Demonstration Test
of PCB-Contaminated Soils, Swanson River, Alaska. OES-910095. 10955 John Jay Hopkins Drive, San
Diego, California
5. Ogden Environmental Services, Inc. 1989. Source Test Data Report for the OES Circulating Bed
Combustor. ACUREX Project No. 9423. OES, Inc., P.O. Box 85178, San Diego, California
6. USEPA 1989. Preliminary Test Burn/Treatability Study of McColl Superfimd Waste in Ogden's
Circulating Bed Combustor Research Facility. Foster Wheeler Enviresponse, Inc. Contract No.
68-03-3255, Risk Reduction Engineering Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency, Cincinnati, Ohio.
'U.S. Government Printing Office: 1992—648-003/40703
59
-------
-------
-------
-------
-------
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
agency
Cincinnati OH 45268-1072
Official Business
Penalty for Private Use, $300
EPA/540/R-92/001
------- |